Police IQ test: what the Jordan case actually decided

0 The short answer

There is no single police IQ test: the agencies whose official pages were read for this article screen applicants with written ability exams, at least one Connecticut town lists the Wonderlic Personnel Test as a step, and the famous "too smart to be a cop" story is a real case, Jordan v. City of New London, in which a Wonderlic score of 33 fell above the range of 20 to 27 that the city interviewed, and a federal district court (1999) and the Second Circuit (2000) held that the cutoff was rational under the equal protection clause, which is a different thing from holding that it was wise or that police work punishes intelligence. The official exam pages read for this article describe tests of reading, writing, reasoning, memory and, in one state, arithmetic, with pass marks, not IQ ceilings. If you are curious about your own level, ACIS offers a normed adult battery as one option among others (Quick at 15 dollars and about 45 minutes, Optimized at 30 dollars and about 110 minutes, Full Scale at 50 dollars and about 175 minutes, prices read October 29, 2026), and the decisive limit is that an online, unsupervised score describes you and does not stand in for any exam a department sets.

1 What is a police IQ test, and do departments give one?

A police IQ test, in the way people use the phrase, is almost never a named intelligence test: it is the written entrance exam a department, state board or civil service agency sets, and in some cases a short cognitive test from a commercial publisher. The searches behind this page (police IQ test, police officer IQ test, cop IQ test, police academy IQ test) mix three different things. The first is the civil service or state exam that ranks or passes applicants, which the next section describes from official pages. The second is a commercial cognitive test such as the Wonderlic Personnel Test, which one Connecticut town lists as a step in its police hiring process on a recruitment page read for this article. The third is an individually administered intelligence battery, which a psychologist may give in a clinical or research setting but which is not what an applicant meets at the entrance stage in any of the official pages opened here.

The Wonderlic deserves one clear sentence of context because the New London story turns on it. It is, in the district court's words, a standardized test of cognitive ability, its manual lists recommended scores for a range of occupations, and it is not a police instrument; the page on what the Wonderlic is today and how its scores relate to IQ covers its format, its publisher's guidance and the limits of converting its scores. This page does not repeat that material. The page on the pre-employment cognitive test covers the wider family of hiring tests, and the page on how hiring tests are used and regulated covers the general law of testing job applicants, which is much broader than one case.

Two cautions keep the rest of the page honest. First, the departments and agencies named here are examples read on specific pages, not a survey of the thousands of agencies in the United States, so nothing below says what "police departments" do as a group. Second, a test used to screen job applicants is a different instrument from a test used to describe a person. An applicant test is built to rank or sort a pool against a job, and its score belongs to that process. A normed battery taken for self knowledge reports where a person stands against adults of the same age. The same number can mean very different things in those two settings, which is the lesson of the court case and the reason the later sections separate them.

2 What do police entrance exams actually test?

Official exam pages describe tests of reading, writing and reasoning, with pass marks or rankings and sometimes arithmetic, memory and spatial tasks, and none of the pages opened for this article describes an upper limit on how well an applicant may score. The pages below were all opened on October 29, 2026. They describe the exam as the agency presents it. They are examples, formats change, and an applicant should always read the current notice of the agency they apply to.

The Oregon State Police describe a Police Officer Selection Test (POST) in a familiarization sheet with four separately timed sections: mathematics (20 questions, 30 minutes, no calculator), reading comprehension (25 questions, 25 minutes), grammar (20 questions, 15 minutes) and report writing (10 questions, 15 minutes). The sheet says that most questions relate to police duty but that no prior knowledge of laws or law enforcement is needed. The California Commission on Peace Officer Standards and Training (POST) describes its entry-level law enforcement test battery as a multiple-choice, fill-in-the-blank written examination designed to measure reading and writing ability, given on paper in 2.5 hours across five sections (spelling, vocabulary, clarity, reading comprehension and a cloze section that the page says was to be removed effective January 1, 2025). That page reports results as a T-score, calls about 50 average, and says that research shows the likelihood of successful academy completion increases for every point above 42.

New York City's Department of Citywide Administrative Services publishes a notice of examination for Police Officer, amended January 14, 2026. The multiple-choice test is taken at a computer, a score of at least 70 percent is required to pass, and the score determines the candidate's place on an eligible list. The notice says the test is designed to assess the extent to which candidates have abilities determined to be important to the tasks of a police officer, and it lists them: deductive reasoning, inductive reasoning, information ordering, memorization, problem sensitivity, spatial orientation, visualization, written comprehension and written expression. The San Diego Police Department's recruiting page describes a multiple-choice written test with three sections (reading comprehension, executive function and written comprehension) and says a candidate must score at least 70 percent to pass. The Town of Windsor, Connecticut lists the Wonderlic Personnel Test as the second step of its police officer hiring process, before physical agility, an oral interview panel with a report writing test, a background investigation, a polygraph examination, a psychological evaluation, a physical exam and drug screen and an interview with the chief. The page gives no passing score for the Wonderlic step and no date, and another list on the same page, headed as the hiring procedure, calls the first test a written aptitude test without naming the Wonderlic.

Reading across the table, two features stand out. The government pages speak about job relevance, pass marks and rankings, and none of them speaks about IQ. And the one commercial cognitive test in the table sits in a hiring sequence of nine steps in which it is the second, followed by seven further steps, among them physical agility, an oral interview, a background investigation and a psychological evaluation. That context matters when reading the New London story, because it concerned a single early screen in one consortium's process.

3 Who was Robert Jordan, and what happened in New London?

Robert Jordan was an applicant for an entry level police officer position in 1996 who scored 33 on the Wonderlic Personnel Test, and the city of New London, Connecticut, did not interview him because it considered only applicants who scored between 20 and 27. Both court opinions state the facts, and they are the best source because each court wrote them down from the record. The Second Circuit's summary order, issued August 23, 2000, says that on March 16, 1996, Jordan and 500 other applicants went through a written screening process run by the Law Enforcement Council of Southeastern Connecticut, Inc., a coalition of fourteen cities and towns, as an initial screen for participating police departments. The testing material included the Wonderlic Personnel Test and Scholastic Level Exam (WPT), which the order says "purports to measure cognitive ability." The district court's ruling, dated August 29, 1999 and filed September 2, 1999, describes the same event in nearly the same terms and adds that the advertisement said the test would generate an eligibility list for distribution to the participating departments. Both texts are reproduced on a page of AELE, a law enforcement legal education center, which is the source read for this section on October 29, 2026.

The score and the guidance around it are the heart of the matter. The users manual for the Wonderlic contained a table of recommended minimum scores for a range of occupations. According to the district court, the suggested minimum for a patrol officer was 22, the manual said an appropriate score for most hiring decisions may run from two points below the suggested score to six points above it, and another Wonderlic document gave a recommended range for a police patrolman of approximately 18 to 30. The Second Circuit's order describes the manual differently and says it listed recommended scores for various professions and cautioned that, because overqualified candidates may soon become bored with unchallenging work and quit, simply hiring the highest scoring employee can be self defeating. It gives Jordan's score as 33, "above the median for any listed occupation," and "well over the normative median of 21 suggested for a police patrol officer." The district court says that, following the recommendation of the manual, New London had decided to consider only applicants who scored between 20 and 27.

How Jordan learned of it is told in both texts. In the fall of 1996 he learned that New London was interviewing candidates and asked about an interview with Keith Harrigan, the assistant city manager in charge of personnel. The Second Circuit says Harrigan told him he would not be interviewed because he did not "fit the profile," that Jordan, who was 46 years old, suspected age discrimination and filed an administrative complaint with the Connecticut Commission on Human Rights and Opportunities, and that the city answered that it had removed him from consideration because he scored 33 and that, to prevent frequent job turnover caused by hiring overqualified applicants, it interviewed only candidates who scored between 20 and 27. The district court, citing Jordan's complaint, says Harrigan informed him that he was ineligible because he scored too high on the written test. Jordan then sued the city and Harrigan in federal court, and the claim, in the words of the district court, was a violation of the Equal Protection Clauses of the United States and Connecticut constitutions stemming from the defendants' hiring practice.

Contemporary news coverage added the human side and a conversion that the court did not make. CBS News, in a report from September 10, 1999, quoted Jordan as saying that, philosophically, he found it offensive "to the entire profession of law enforcement" and adding, "We all know talented, intelligent people that pursue successful careers in law enforcement." The same report says he had served as a part-time officer in Groton Long Point, Connecticut, in 1989, as a seasonal officer for the state Department of Environmental Protection from 1993, and that he was then a corrections officer for the state. ABC News, on September 8, 2000, reported that he had lost the appeal, that he said he did not plan further legal action, and that he had worked as a prison guard since taking the test. Both outlets wrote that 33 was the equivalent of an IQ of 125 and that the average score for police officers was 21 or 22, the equivalent of an IQ of 104. Those equivalents are the outlets' own conversions. The courts do not give them, and the page on how a Wonderlic score relates to IQ explains why a conversion from a short screening score carries a wide band. On a normal curve with a mean of 100 and a standard deviation of 15, an IQ of 125 sits near the 95th percentile and an IQ of 104 near the 61st, which is our arithmetic and not the outlets', and the IQ percentile calculator shows the same conversion for any score.

4 What did the district court decide, and on what legal standard?

The United States District Court for the District of Connecticut, Judge Peter C. Dorsey, granted summary judgment to the city on August 29, 1999, holding under rational basis review that the cutoff did not deny Jordan equal protection, and writing that he "may have been disqualified unwisely but he was not denied equal protection." The case number was Civil No. 3:97CV1012 (PCD), reported as 1999 U.S. Dist. LEXIS 14289 and 15 IER Cases (BNA) 919. Summary judgment is a ruling without a trial: the opinion explains that it is confined to instances where there is no genuine issue as to any material fact and the moving party is entitled to judgment as a matter of law. The ruling was therefore about law applied to a record, and not a finding after a full trial about whether high scorers make worse officers.

The legal standard did most of the work. The court wrote that Jordan conceded that he was not a member of a suspect class and that there is no fundamental right to employment as a police officer, so rational basis review was the proper standard. Under that standard a government classification is upheld if it is rationally related to a legitimate purpose. Jordan also conceded that increasing employment longevity and reducing the high costs of rapid employee turnover are legitimate government purposes, and that limiting an applicant pool to a manageable level is a legitimate goal. That left a single question, which the court stated this way: whether the defendants' means were rationally related to those goals. The court quoted the rule that there is no constitutional violation if there is any conceivable state of facts that would provide a rational basis for the government's conduct, and also the limit from the Supreme Court's Cleburne decision that a state may not rely on a classification whose relationship to an asserted goal is so attenuated as to render the distinction arbitrary or irrational.

On the evidence, Jordan did not lose for lack of argument. The ruling says he provided evidence of a lack of definitive relationship between high cognitive ability and job satisfaction, performance and longevity, and argued that data suggesting such a correlation conflicts with current thinking in industrial psychology circles. The court's answer was that the defendant "need show only that there was reason to believe that employing the classification could be beneficial in achieving their stated goal," and that it is not the function of courts to judge the wisdom of particular business policies but to ensure that they are made on a rational basis. It found two reasons the city could rationally rely on. The defendants followed the instructions accompanying the Wonderlic test, on which they rationally could have relied as authoritative, and "a body of professional literature concludes that hiring overqualified applicants leads to subsequent job dissatisfaction and turnover. While such studies have been challenged, it cannot be said that their conclusions have been refuted."

The court also rejected a second argument, that the city used the scores as an afterthought to reduce the number of applicants. It wrote that defendants reasonably may apply the publisher's objective criteria to screen an otherwise unmanageable number of applicants, that they need not have decided on the numbers before the flood of applications arrived, and that the fact they could have ignored the guidelines does not make the decision to follow them arbitrary. Its closing logic was that it is not for a trier to decide between two bodies of evidence supporting and contradicting the validity of a basis. The judgment was entered on September 2, 1999, ordering that judgment be entered for the defendants and the case closed. Notice what the ruling does not say. It does not say that scoring 33 predicts anything about how Jordan would have performed, it does not endorse the Wonderlic manual's advice as sound psychology, and it does not say the evidence on the question was settled. It says the city had a rational basis, and that is all the standard requires.

5 What did the Second Circuit decide in 2000?

The United States Court of Appeals for the Second Circuit, in a summary order dated August 23, 2000 (No. 99-9188, 2000 U.S. App. LEXIS 22195), affirmed, holding that New London's use of an "upper cut" did not violate the equal protection clause. The panel was Circuit Judges Jon O. Newman and Rosemary S. Pooler and District Judge Lloyd D. George of the District of Nevada, sitting by designation. The document carries a notice that the rules of the Second Circuit may limit citation to unpublished opinions, so it is a short order and not a published precedent, and a reader should treat it as the appeal court's decision in this one case and not as a rule for all cities. The panel reviewed the grant of summary judgment de novo, meaning it looked at the question afresh.

The order repeats the standard in plain terms. Because Jordan conceded that there is no fundamental right to employment as a police officer, a classification that neither proceeds along suspect lines nor infringes fundamental rights "must be upheld against equal protection challenge if there is any reasonably conceivable state of facts that could provide a rational basis for the classification." The panel added that it requires no supporting empirical evidence, and it recalled the Supreme Court's view that classifications need not be perfect because the problems of government are practical ones. Then it applied "that lenient standard of review," and wrote that even absent a strong proven statistical correlation between high scores on the Wonderlic test and turnover resulting from lack of job satisfaction, it is enough that the city believed, on the basis of material prepared by the test maker and a letter along similar lines sent by the Law Enforcement Council, that there was such a connection.

Two sentences carry the holding. The first deals with Jordan's evidence: he "presented some evidence that high scorers do not actually experience more job dissatisfaction, but that evidence does not create a factual issue, because it matters not whether the city's decision was correct so long as it was rational." The second states the policy rationale the panel accepted: "Even if unwise, the upper cut was a rational policy instituted to reduce job turnover and thereby lessen the economic cost involved in hiring and training police officers who do not remain long enough to justify the expense." The panel said the city could rationally have relied on the test maker's guide as justification for reducing the size of the applicant pool "with both a low and a high cut off," and it found Jordan's remaining arguments without merit.

The press reported the result as a ruling that a police department could bar high IQs. ABC News headlined its report "Court OKs Barring High IQs for Cops." That headline compresses a narrower legal point. What the courts approved was a rational basis for one city's cutoff, tested under the lenient standard the Second Circuit itself named, with a conceded absence of a suspect class and of a fundamental right. They did not settle the psychology, and the Second Circuit said so in effect when it wrote that correctness does not matter so long as the policy is rational. That is the reason this page keeps the legal question and the scientific question in separate sections, and why the research on whether cognitive ability helps or hurts police performance is reviewed later with its own sources.

6 What did the case not decide, and is there evidence of a score ceiling today?

The case decided one narrow legal question about one city's 1996 screening range, and nothing in the opinions or in the official exam pages opened for this article shows that police departments in general reject high scorers. It is worth listing what the rulings leave open, because the story is usually told as though they settled more. They did not find that high scorers perform worse as officers; the district court expressly declined to choose between the two bodies of evidence. They did not find that New London's range was good policy; both courts used the word unwise as a possibility they did not need to resolve. They did not say the Wonderlic is a valid selector for police work, and they did not address other cities, other tests or other claims. The lawsuit as both opinions describe it rested on equal protection. Jordan's suspicion of age discrimination led to an administrative complaint, but it is not the claim the courts decided.

Do any departments use a ceiling today? No source opened here says so. The agency pages read on October 29, 2026 describe pass marks (70 percent in New York City and San Diego, a T-score scale in California) and, in New York, a rule that the score determines the candidate's place on an eligible list. The Windsor, Connecticut, page lists the Wonderlic Personnel Test as a step without printing a passing score or any range, so it neither confirms nor excludes a cutoff of any kind. A reader who wants to know what a specific agency does has to ask that agency. The honest summary is that a 1996 range is documented in two court opinions, and that no page opened for this article documents the same practice today.

The comparison with military testing is instructive, because the armed forces also sort applicants with a cognitive screen, but on a different logic. As the page on military IQ testing reports from the official ASVAB program, the services use a percentile score to decide whether an applicant is eligible to enlist, the minimum varies by branch, and separate composite scores assign applicants to jobs. The page comparing the IQ and ASVAB explains how an enlistment score relates to an IQ. The contrast shows that a cognitive screen is a design choice: the same kind of test can be used as a floor, a ranking or a band, and the consequences for a high scorer differ in each case. New London used a band.

7 Does cognitive ability predict police performance?

In the law enforcement research read here, cognitive ability tests show higher validities for performance in training than for performance on the job, the job validities are modest, and no study read here reports that higher scorers do worse. The most direct source is a validity generalization study by Hirsh, Northrop and Schmidt in Personnel Psychology (1986), which applied the Schmidt-Hunter procedure to validity data for cognitive ability tests for law enforcement occupations. Its abstract reports that for studies using performance in training programs as the criterion, validities ranged from .41 to .71. For studies using performance on the job, validities were between .17 and .31 for three test types and, for the remaining test types, estimated mean true validities ranged from .10 to .26. A validity coefficient is a correlation between test scores and a criterion, so a value of .2 is a modest relationship that leaves most differences between officers unexplained. The authors discussed possible reasons for the lower job figures, including possible low validity of the criterion because supervisors lack the opportunity to observe behavior, and the potential role of noncognitive factors in law enforcement success.

Later single studies point the same way. De Meijer, Born, Terlouw and Van Der Molen studied Dutch police selection in the International Journal of Selection and Assessment (2008) with 2,365 majority and 682 minority applicants and wrote that their findings confirmed the relatively low predictive validity of cognitive ability generally found for police jobs. For the minority group, training performance appeared to be mainly predicted by the cognitive ability test. Henson, Reyns, Klahm and Frank, in Police Quarterly (2010), analyzed personnel data on 486 officers hired between 1996 and 2006 by a Midwestern police department and reported that recruits scoring higher on the civil service exam consistently performed better on multiple academy outcome measures, that those with better overall academy results generally received better evaluations from superiors, and that higher education was not related to any of the success measures. The two studies cover different countries, different decades and different criteria, and neither is a measure of the Wonderlic, so they support a cautious summary: tests like these help forecast performance in training and the academy, and forecast later field performance less well.

None of these studies, in the abstracts read for this page, reports a penalty at the top. The most relevant general test of that idea is by Arneson, Sackett and Beatty in Psychological Science (2011), who examined the relationship between cognitive ability and performance across the whole score range in four large-scale data sets from education and employment. Their abstract says monotonicity was maintained in all instances and that, contrary to the good-enough hypothesis, the relationship was commonly stronger at the top end of the score distribution than at the bottom end. Those data sets were not police samples, so this page does not claim the result for officers. In those four data sets the relationship did not flatten at the top, but one paper cannot settle the question for police work or for any other single job. For the broader literature on ability and work outcomes, the page on IQ and job performance covers the meta-analyses, and the page on reliability and validity explains what a validity coefficient can and cannot say about one person.

8 Can a score be too high for a job?

The version of "too high" that has research behind it in the sources read here is overqualification, and that literature concerns how people feel about a job and what they do afterwards, which is related to but not the same thing as a test score. The turnover logic behind New London's cutoff is not absurd on its face: a person who finds work unchallenging may leave sooner, and replacing an officer after academy training is costly, as the Second Circuit noted. The relevant meta-analysis is by Harari, Manapragada and Viswesvaran in the Journal of Vocational Behavior (2017), titled as a meta-analysis of perceived overqualification. The word perceived matters: the construct is an employee's own sense of being overqualified. A university news release about the study says employees who feel overqualified tend to be less satisfied with their jobs and less committed to their organizations, and it does not mention turnover. This page cannot extend that to police officers or to test scores, because the release and the title do not do so.

That gap is the honest answer to the "too high" question. A pattern in which people who feel overqualified report lower satisfaction does not show that a Wonderlic score of 33 predicts a short career, and the district court itself noted that the studies linking overqualified hiring to turnover had been challenged without being refuted. The court did not have to resolve it, and an applicant, a department or a reader should not resolve it by assumption either. Cognitive ability also relates to many outcomes at once. The page on how much IQ predicts success sets the numbers in context, and the page on average IQ estimates by profession states that every profession contains a wide spread of people, which is a reminder that a single band cannot describe a job. Whether a given cutoff reduces turnover is an empirical question about a particular department, and the professional Standards discussed below describe the kind of evidence that would answer it.

9 Which abilities do police exams sample, in CHC terms?

The abilities that the official exam pages name map onto a handful of broad abilities in the Cattell-Horn-Carroll (CHC) model, mainly comprehension-knowledge (Gc), fluid reasoning (Gf), memory (working memory, Gwm, and learning and retrieval, Glr), visual-spatial processing (Gv) and, where arithmetic appears, quantitative knowledge (Gq). That mapping is our reading of the agency descriptions and not a statement by any agency, and it is approximate, because agencies write their abilities for the job and CHC writes them for the structure of cognition. New York City's notice defines deductive reasoning as applying general rules to specific problems and inductive reasoning as combining separate pieces of information to form general rules, which are the two classic reasoning abilities placed under Gf. It defines memorization as remembering words, numbers, pictures and procedures, and visualization as imagining how something would look when it is moved around or rearranged, as in unfolding or rotation, which correspond to memory abilities and to Gv. Written comprehension and written expression involve English reading and writing, which CHC treats as reading and writing ability and which, in an IQ battery, shows up through vocabulary and verbal knowledge (Gc). The Oregon sheet adds mathematics without a calculator, a quantitative demand (Gq).

ACIS tests broad abilities with these names, and the subtest pages describe the formats. The Vocabulary page describes how precisely a person understands words, a classic indicator of crystallized knowledge. The Matrix Reasoning page describes inferring a missing element from a rule structure, which it discusses under fluid reasoning and inductive reasoning. The Visual Puzzles page describes mentally combining separate pieces into a whole, discussed under spatial visualization. The Digit Span page describes holding and reordering items in immediate awareness, working memory. The page on the CHC model lists the broad and narrow abilities behind these terms. Two cautions apply. A shared name does not make two tasks equivalent, since format, timing and content differ, so no ACIS subtest stands in for an item on any agency exam. And the exams in the previous section are job screens built around the tasks of an officer, with the ability names drawn from that analysis, so their scores answer a different question than a profile of broad abilities does.

10 How should you read an entrance exam score or an IQ score?

A pass mark, a place on a list and an IQ are three different kinds of number, and the Standards for Educational and Psychological Testing ask that cut scores, the use of a test for selection and the evidence behind that use be documented. A pass mark such as 70 percent is a criterion: it says whether an applicant cleared a bar the agency set. A place on an eligible list is a rank: it says where an applicant stands among the others who sat the exam and depends on who else applied. A T-score, as California's page uses it, and an IQ are norm referenced: they say where a person stands against a reference group. The page on how IQ scores are normed explains that logic and the page on what an IQ score means covers the bands. Percentiles and standard scores are not interchangeable, and the page on the IQ standard deviation of 15 shows why a given gap in points means different things at different places on the curve.

The Standards (AERA, APA and NCME, 2014), the joint publication of the American Educational Research Association, the American Psychological Association and the National Council on Measurement in Education, address cut scores directly. In the open access edition, Standard 5.21 says that when score interpretations involve cut scores, the rationale and procedures used for establishing them should be documented clearly. Its comment notes that cut scores may be set to select a specified number of examinees, for example to identify a fixed number of job applicants for further screening, in which case little further documentation may be needed about how they were set, although attention should be paid to the rationale for using the test in selection and to the precision of comparisons among examinees. The comment to Standard 5.23 adds that in employment settings where test scores have been shown to relate to job performance, the precise relation between test and criterion may have little bearing on a cut score chosen to yield a predetermined number of candidates. Both comments concern cut scores chosen to select a number of candidates when scores are known to relate to job performance. They do not mention a ceiling that excludes high scorers, they are not a legal test, and they do not say that New London's range met or failed any professional standard. They show only that a cut chosen to limit a pool is documented differently from a cut that classifies people into categories with distinct meanings.

Chapter 11 of the Standards, on workplace testing, also reminds users that performance can be hindered by a source of variance unrelated to the construct of interest, with the example of a person unfamiliar with computer technology facing the user interface of a simulation, and it says sound practice involves monitoring the assessment process to prevent undue disadvantages or advantages caused by such factors. That is one reason the page on whether IQ tests are biased matters for any timed, unsupervised or computer delivered test. A single score is also an estimate with error, and the 95 percent confidence interval around a Full Scale IQ is the honest way to read it.

11 What can a curious adult take, and what about practice questions?

An adult curious about their own level can take a normed adult battery online or ask a psychologist for an individually administered one, and a person preparing for a specific police exam should rely on the agency's own materials, because research shows that scores rise on retest and rise more with coaching and identical forms. For the first goal the options are laid out on the pages that cover them: where to take an IQ test compares every route, in person and online, and the best online IQ tests ranks the field. The pages on the free IQ test, an IQ test for adults and a cognitive ability test describe the entry points on this site. For readers whose process uses another employer test, the siblings of this page describe them from their publishers' pages: the CCAT, the critical thinking test and the General Aptitude Test Battery.

The search for police IQ test questions deserves a plain answer. This page does not print items from any exam, and it links to no site that sells practice tests. Agencies sometimes publish their own familiarization material, as the Oregon State Police do for the POST, and that is the right place to see an exam's format and timing. On whether practice helps, Hausknecht, Halpert, Di Paolo and Moriarty Gerrard summarized 50 studies of practice effects for tests of cognitive ability in the Journal of Applied Psychology (2007). Their abstract reports results from 107 samples and 134,436 participants with an adjusted overall effect size of .26, with larger effects when practice was accompanied by test coaching and when identical forms were used. So scores tend to rise on retest, and more so with coaching and identical forms, which is a reason to learn the format of the actual exam from the agency itself.

Practical conditions matter for any unsupervised session. Use a quiet room and a device you are comfortable with. Digit Span, Alphanumeric Sequencing and Arithmetic present their items as audio, so working sound is required for them. Breaks are allowed between subtests, progress is saved, and the assessment has 30 days to complete, so a long form does not need to be done in one sitting.

12 Where ACIS sits, and what a buyer should do

ACIS is an online, unsupervised, adult normed battery for self knowledge, and it is not a police exam, not a hiring tool and not a substitute for a clinical evaluation. The site's research page states the limit literally: ACIS "is not intended for clinical diagnosis, hiring decisions, accommodation determinations or admission to high IQ societies." ACIS makes no claim about any agency's exam, and nothing on this page suggests that a report predicts, prepares for or equals one. The English language is required, the manual says verbal and knowledge tasks are affected by English proficiency and educational exposure, and the session is unsupervised, so the score reflects the conditions in which you took it. The Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014) are the right yardstick for that judgment: Standard 11.12 asks, when the use of a test for personnel selection relies on a link between the construct the test represents and job performance, for evidence that the scores are reliable, that the content adequately samples that construct, and that the construct relates to major factors of the job performance domain. ACIS makes no claim of that kind, which is why it is not a selection tool.

What ACIS does offer is a profile. It has 20 subtests in six domains and reports a Full Scale IQ with its 95 percent confidence interval and six index scores (verbal comprehension, fluid reasoning, quantitative reasoning, visual spatial, working memory and processing speed), on the standard scale with a mean of 100 and a standard deviation of 15, with adult norms for ages 16 to 90. It is a one time payment with no subscription, and the prices read on October 29, 2026 can change: Quick is 15 dollars for 6 subtests in 3 domains and about 45 minutes, Optimized is 30 dollars for 13 subtests in 5 domains and about 110 minutes, and Full Scale is 50 dollars for all 20 subtests and about 175 minutes. The Custom Test costs 5 dollars per subtest. There is a free trial with no card, one per person. The guarantee reads: "Full refund within 5 days if the report does not deliver the features described at checkout or a technical issue prevents you from accessing your results." The technical manual documents the instrument for readers who want the evidence.

The next step depends on the reader. If you are applying to a department, read that agency's current notice of examination and use its own familiarization material, and treat the Jordan case as a historical example, not a forecast. If you are curious about your own profile, start with the free trial or an example report, take the form that fits your question, and read the report as an estimate with a band around it. If a decision with consequences hangs on the number, see a licensed psychologist.

13 Sources Behind This Page

Every figure and quotation above comes from one of the following sources, each linked where it is first used and each opened or verified on October 29, 2026. The court texts were read in full as reproduced by AELE. The percentile equivalents for IQ 125 and IQ 104 and the 85 minute total of the Oregon sections are our own arithmetic, labeled as such in the text. The CHC mapping of exam abilities is also our reading. Research results are group findings and describe no individual officer or applicant.

• AERA, APA and NCME. Standards for Educational and Psychological Testing. Washington, DC: American Educational Research Association, 2014 (Standards 5.21, 5.23 and 11.12 and the chapter 11 background).
• ABC News. Court OKs Barring High IQs for Cops. September 8, 2000. Read October 29, 2026.
• AELE. Jordan v. City of New London, No. 99-9188, 2000 U.S. App. LEXIS 22195 (2d Cir. Aug. 23, 2000) (summary order), and Jordan v. City of New London, Civ. No. 3:97CV1012 (PCD), 1999 U.S. Dist. LEXIS 14289 (D. Conn. Aug. 29, 1999). Full text of both decisions as hosted by AELE. Read October 29, 2026.
• Arneson J, Sackett P and Beatty A. Ability-Performance Relationships in Education and Employment Settings. Psychological Science, 2011, volume 22, issue 10, pages 1336 to 1342.
• California Commission on Peace Officer Standards and Training. LE Entry-Level Test Battery Applicant FAQs. Read October 29, 2026.
• CBS News. Too Smart To Be A Cop? September 10, 1999. Read October 29, 2026.
• City of New York, Department of Citywide Administrative Services. Notice of Examination, Police Officer, Exam Nos. 6314, 6315, 6316, 6323, 6318, 6319, 6320, 6321 and 6322. Amended January 14, 2026. Read October 29, 2026.
• De Meijer L, Born M, Terlouw G and Van Der Molen H. Criterion-Related Validity of Dutch Police-Selection Measures and Differences Between Ethnic Groups. International Journal of Selection and Assessment, 2008, volume 16, issue 4, pages 321 to 332.
• EurekAlert. News release on the meta-analysis of perceived overqualification by Harari and colleagues. Read October 29, 2026.
• Harari M, Manapragada A and Viswesvaran C. Who thinks they're a big fish in a small pond and why does it matter? A meta-analysis of perceived overqualification. Journal of Vocational Behavior, 2017, volume 102, pages 28 to 47.
• Hausknecht J, Halpert J, Di Paolo N and Moriarty Gerrard M. Retesting in selection: A meta-analysis of coaching and practice effects for tests of cognitive ability. Journal of Applied Psychology, 2007, volume 92, issue 2, pages 373 to 385.
• Henson B, Reyns B, Klahm C and Frank J. Do Good Recruits Make Good Cops? Problems Predicting and Measuring Academy and Street-Level Success. Police Quarterly, 2010, volume 13, issue 1, pages 5 to 26.
• Hirsh H, Northrop L and Schmidt F. Validity generalization results for law enforcement occupations. Personnel Psychology, 1986, volume 39, issue 2, pages 399 to 420.
• Oregon State Police. Police Officer Selection Test (POST) Exam Familiarization. Read October 29, 2026.
• San Diego Police Department. Written Test, recruiting page. Read October 29, 2026.
• Town of Windsor, Connecticut. Police Recruitment. No date shown on the page. Read October 29, 2026.

14 Frequently Asked Questions