A free 54-question practice simulation of all three batteries: verbal, quantitative, and nonverbal. Instant estimated percentile and stanine, plus a worked answer for every question. Untimed, so your child can think.
Take CogAT TestA note for parents: this practice test is untimed by design. Sit with your child, read questions aloud if that helps, and treat it as a game rather than an exam. The goal is familiarity and confidence, not a number.
Everything below is free and appears the moment your child answers the last question.
Not a worksheet of sample questions: a full three-battery run that ends in an estimated SAS score, a percentile, a stanine, and a bar for each battery. The whole report is free.
Every question ships with a short explanation that names the reasoning skill it trains, written so you can sit with your child and talk through each one.
Untimed, big tap targets, one question at a time, and no clock on the screen. Retakes are tracked on your device, so improvement is visible.
A working miniature of the free report, filled with example data. Click the tabs to look around. Yours is generated instantly from your child's own answers.
Sample report · example data
Average
Estimated SAS score (mean 100, SD 16) with stanine 7. This is a practice-level estimate from one untimed attempt, not an official CogAT score.
An estimated SAS score of 114 sits at the 82nd percentile, stanine 7.
Breakdown
Reasoning with words: analogies, sentence completion, and sorting words into groups.
Estimated 83rd percentile for this battery
Reasoning with numbers: number analogies, number series, and balance puzzles.
Estimated 89th percentile for this battery
Reasoning with shapes and pictures: figure matrices, classification, and pattern series.
Estimated 50th percentile for this battery
Strongest battery
Quantitative Battery
15 of 18 correct. Reasoning with numbers: number analogies, number series, and balance puzzles.
Most room to grow
Nonverbal Battery
11 of 18 correct. Reading the worked answers together is the fastest way to improve here.
Gifted screening
Many district gifted programs screen around the 95th to 98th percentile, though every district sets its own bar and most weigh more than one measure. This practice attempt lands at the 82nd percentile overall (estimated), with the strongest battery, Quantitative Battery, at the 89th. Read it as a picture of where your child is comfortable today, not as a prediction of any screening decision.
83rd
Verbal Battery (stanine 7)
89th
Quantitative Battery (stanine 8)
50th
Nonverbal Battery (stanine 5)
One practice attempt is not a qualification decision. Official CogAT scores come from a proctored administration with age norms, and districts combine them with teacher input and other evidence.
Answer review
A short worked explanation for every question, free, written to be read together. Mistakes are listed first.
Why: Every picture in the grid is drawn as an outline with an empty middle. The outline diamond is the only answer that is not filled in, so it belongs with the group.
Skill trained: figure classification
Why: Both example pairs multiply by 3: 4 becomes 12, and 2 becomes 6. So 5 times 3 makes 15. The answer 13 is the trap for adding 8, which fits the first pair but not the second.
Skill trained: number analogy
Why: Minutes are the small parts that make up an hour, and days are the small parts that make up a week. This is a part-to-whole analogy.
Skill trained: verbal analogy (part to whole)
Three rows from the sample. Your report lists all 54 questions, with the mistakes first.
Improve
Example history. Attempts are saved on your own device, so a retake shows exactly how much your child improved.
Example data, not a real child's result. Buttons inside the sample are part of the mock. All scores are practice-level estimates, not official CogAT scores.
Find 25 quiet minutes. Read questions aloud if your child prefers; the real primary-level CogAT is read to children too.
Eighteen questions per battery, rotating between words, numbers, and pictures so no single format tires your child out.
An estimated SAS score, percentile, and stanine, plus a bar for each battery showing where your child is most comfortable.
Open any question, read the worked explanation together, and treat mistakes as puzzles to solve rather than errors to fix.
Three formats per battery. Read these together before starting and none of the 54 questions will feel unfamiliar.
Prints these nine worked samples on paper, free, no download or email needed.
Two words connect in some way; your child finds the word that connects to a third word in the same way. The youngest levels of the official test use pictures instead of words.
Worked sample
Puppy is to dog as kitten is to:
A. bone B. cat C. collar D. bark
Answer: B. cat
A puppy grows into a dog, and a kitten grows into a cat. The skill is spotting the relationship first and only then looking at the choices.
A sentence has a missing word, and only one choice makes the whole sentence make sense. This rewards careful listening and reading rather than vocabulary depth.
Worked sample
The snow melted because the sun was so ____.
A. cold B. bright C. warm D. high
Answer: C. warm
The word "because" asks for the reason snow melts, and warmth is what melts snow. Bright is the trap: sunny days are bright, but brightness is not what melts snow.
Three words form a group that shares something; your child picks the answer that belongs with them (or, in the flipped version, the word that does not belong).
Worked sample
Red, blue, yellow. Which word belongs with this group?
A. crayon B. green C. rainbow D. paint
Answer: B. green
Red, blue, and yellow are all colors, so the group needs another color. Crayon and paint are things that HAVE colors, which is a different kind of connection.
One pair of numbers follows a hidden rule; your child applies the same rule to a new number. The rules stay simple at this level: add, subtract, double, halve.
Worked sample
2 is to 6, and 1 is to 3. So 3 is to what number?
A. 7 B. 5 C. 9 D. 12
Answer: C. 9
Both example pairs multiply by 3: 2 becomes 6, and 1 becomes 3. So 3 times 3 makes 9. The answer 7 is the trap for adding 4, which fits the first pair but not the second.
A row of numbers follows a pattern and your child finds the next one. The key habit is looking at the jumps between numbers, not the numbers themselves.
Worked sample
What number comes next: 20, 18, 16, 14, ?
A. 13 B. 12 C. 10 D. 15
Answer: B. 12
The numbers count down by 2 each time, so after 14 comes 12. Series that go down catch out children who assume patterns always grow.
An equation or balance has a missing number. These are gentle first steps toward algebra: the child reasons about what value makes both sides equal.
Worked sample
Which number goes in the blank: 4 + __ = 9?
A. 6 B. 4 C. 13 D. 5
Answer: D. 5
The blank is the number that carries 4 up to 9, and 4 + 5 = 9. The answer 13 is the trap for children who add the two numbers they can see.
A grid of shapes follows rules across its rows and columns, with one cell missing. Your child works out the rules and picks the shape that completes the grid.
Worked sample
Each row keeps its shape and each column keeps its color. Which picture goes in the empty square?
Four candidate shapes, one of which fits both the row rule and the column rule.
Answer: The shape matching its row with the color matching its column.
Matrices train rule-finding: hold the row rule and the column rule in mind at the same time, then check each choice against both.
A set of shapes all share one property, like "all circles" or "all striped". Your child picks the answer shape that shares it too.
Worked sample
All the shapes in the group have four sides. Which answer belongs with them?
A diamond, a circle, a star, and a triangle.
Answer: The diamond, because it also has four sides.
The skill is finding what the group has in common while ignoring what varies. Here color and size change freely; only the number of sides matters.
A row of shapes repeats or changes step by step, and your child picks what comes next. Our simulation uses this format in place of the official paper-folding items, which need physical paper to practice honestly.
Worked sample
The pattern goes: big, small, big, small, big. Which comes next?
A big shape or a small shape.
Answer: The small shape, continuing the alternation.
Series items train sequence thinking: say the pattern out loud ("big, small, big, small...") and the next step usually announces itself.
The Cognitive Abilities Test, published by Riverside Insights, is the reasoning test American schools use most often to screen children for gifted and talented programs. It is not a school achievement test: it does not ask what your child has been taught. Instead it measures how your child reasons with words, with numbers, and with shapes, which is why the results are organized into three separate batteries: Verbal, Quantitative, and Nonverbal.
Schools currently administer Form 7 or the newer Form 8. Both use the same three-battery structure and the same nine question formats, and both are group-administered, usually on a computer or with paper booklets in the classroom. Depending on the school, a child may sit the full form (all three batteries, roughly 90 minutes to two hours split over sessions) or a shorter screening form of about 30 minutes.
The test comes in levels named for the age of the child, from Level 5/6 for kindergarten up to Level 17/18 for the last years of high school. At the youngest levels (5/6 through 8) the questions are read aloud and use pictures rather than written words, so a child does not need to be a fluent reader to take it. From about Level 9 upward, the child reads independently.
| Level | Typical grade | How it is presented |
|---|---|---|
| 5/6 | Kindergarten | Read aloud, picture-based |
| 7 | Grade 1 | Read aloud, picture-based |
| 8 | Grade 2 | Read aloud, transitioning to text |
| 9 | Grade 3 | Child reads independently |
| 10 | Grade 4 | Child reads independently |
| 11 to 12 | Grades 5 to 6 | Child reads independently |
| 13/14 to 17/18 | Grades 7 to 12 | Child reads independently |
Where this practice test fits: our 54 questions target the Grade 2 to 4 screening window (Levels 8 to 10), which is when most districts run their universal gifted screening. The stems are written for a 7 to 9 year old to read alone. For a kindergartner or first grader, read every question aloud and treat the score loosely, since the real test at that age is fully picture-based. For a fifth grader or older child, expect the questions to feel easy; a strong score is less informative than it would be for a younger child, though the answer review is still useful practice.
One more honest note: this is a practice test, written from the publicly documented structure of the test. It is not affiliated with Riverside Insights, it does not contain real CogAT items, and no practice score predicts an official one. What it can do is make the question formats familiar, show you where your child is comfortable, and give you something concrete to talk about together. Our methodology page explains how we build and score our tests, and who we are.
CogAT score reports look intimidating because one performance is expressed four different ways. The chain is simpler than it looks. First comes the raw score: how many questions your child answered correctly. The raw score is converted to a Universal Scale Score (USS), a common yardstick that lets the publisher compare performances across different levels and forms of the test.
The USS is then compared against other children the same age, producing the number most districts actually use: the Standard Age Score (SAS). The SAS scale has a mean of 100 and a standard deviation of 16, with a maximum of 160. An SAS of 100 means perfectly typical reasoning for the child's age; 116 is one standard deviation above; 132 is two. From the SAS come the last two numbers: the percentile rank (the share of same-age children scoring lower) and the stanine, which compresses everything onto a 1-to-9 scale that teachers use for quick grouping.
| SAS score | Percentile | Stanine | Common label |
|---|---|---|---|
| 132 and above | 98th and above | 9 | Very high |
| 124 to 131 | 93rd to 97th | 8 | Well above average |
| 116 to 123 | 84th to 92nd | 7 | Above average |
| 108 to 115 | 69th to 83rd | 6 | High average |
| 93 to 107 | 32nd to 68th | 5 | Average |
| 84 to 92 | 16th to 31st | 3 to 4 | Low average |
| Below 84 | Below 16th | 1 to 3 | Below average |
Your child receives these numbers for each battery separately and for the composite. The battery profile often matters more than the composite: a child with a 97th percentile Quantitative score and a 60th percentile Verbal score is a different learner from a child with three 85s, even if their composites look similar. Our practice report mirrors this structure, with an estimated SAS score, percentile, and stanine for the attempt and a separate bar for each battery, all labeled as estimates because a free practice test has no standardization sample behind it.
There is no national cutoff, and any resource that names one number is simplifying. Gifted identification in the United States is decided district by district, and the same SAS score that qualifies a child in one town would not in the neighboring one. That said, the patterns are directional, and knowing them helps you read your district's policy.
| District approach | Typical bar | Notes |
|---|---|---|
| Selective full-time gifted programs | 98th percentile composite (SAS around 132) | Often combined with achievement scores and teacher rating scales |
| Broad enrichment or pull-out programs | 90th to 95th percentile | Sometimes on the composite, sometimes any single battery |
| Single-subject acceleration | High score in the matching battery | A 97th percentile Quantitative score can matter more than the composite for math placement |
| Universal screening pass | Top 10 to 20% advance to a second round | The CogAT is the first filter, not the final decision |
Three caveats belong next to any table like this. First, many districts weigh multiple measures: teacher checklists, portfolios, achievement tests, and sometimes a second ability test. Second, some districts use local norms (comparing your child to their own district's students) or building norms rather than national ones, which shifts every cutoff. Third, a single test on a single morning measures a child with a cold, a skipped breakfast, or a fire drill in the middle of it. Districts know this, which is why most have appeal and retest procedures.
The only reliable way to know your district's bar is to ask for its written identification policy, which public districts must generally share. And it bears repeating for this page specifically: one free practice attempt, taken untimed at home, predicts nothing about a screening decision. It tells you which formats are familiar and which batteries feel comfortable, and that is all we claim for it.
Parents researching gifted testing meet a soup of acronyms, and the differences matter because the tests measure different things. The short version: CogAT, NNAT, and OLSAT are reasoning (ability) tests that districts use for gifted screening, while the ITBS is an achievement test that measures what a child has already learned in school. A district may use any of them, and some pair an ability test with an achievement test deliberately.
| Test | Publisher | What it measures | Format | Where you meet it |
|---|---|---|---|---|
| CogAT | Riverside Insights | Verbal, quantitative, and nonverbal reasoning | 3 batteries, 9 question types, group-administered | The most widely used gifted screener in US public schools |
| NNAT | Pearson | Nonverbal reasoning only (figures and patterns) | One section of figural items | Districts with many multilingual learners; language-free by design |
| OLSAT | Pearson | Verbal and nonverbal school ability | 2 sections | Historically the New York City G&T test; still used regionally |
| ITBS | Riverside Insights | Achievement: reading, language, math taught in school | Subject tests | Paired with the CogAT in some districts to compare ability with achievement |
Two practical implications. If your district uses the NNAT, nonverbal practice matters most, and verbal polish matters little: the test has no words at all. If your district pairs the CogAT with the ITBS, the district is often looking for the gap between ability and achievement, so the profile across both tests, not either alone, drives the decision. Our practice simulation covers the CogAT's three-battery structure, which also happens to give reasonable format exposure for the OLSAT's verbal and nonverbal sections and the NNAT's figural items.
There is a real ethical line in test preparation, and it is worth naming plainly. On one side sits familiarity: making sure a seven year old understands what an analogy question is asking before the first time one counts. On the other side sits coaching designed to inflate a score past a child's actual reasoning ability, which, when it works, places a child in a program that will feel like a permanent uphill walk. The first helps your child; the second sets them up.
Familiarity is legitimate because the test assumes it. Test authors expect children to understand the question formats; a child who has never seen a figure matrix spends their mental energy decoding the instructions instead of reasoning. One or two relaxed practice sessions close that gap completely, and the score that emerges is a fairer measure of the child, not a distorted one. That is the entire ambition of this page.
What ethical preparation looks like: a couple of short, playful sessions in the weeks before testing, not nightly drills. Read the worked explanations together and let your child explain answers back to you, since explaining is where the reasoning consolidates. Frame the real test as "puzzles at school", keep your own anxiety out of the room, and protect sleep the week of testing. Pattern-rich play (blocks, card games, "which one does not belong" games in the car) builds the same muscles without a screen.
What to avoid: months-long bootcamps, memorizing leaked or "real" questions (which is both against test security rules and self-defeating), and any framing where the child feels the family's hopes riding on a morning's work. If a preparation program promises a specific percentile, close the tab. And if your child scores below a cutoff, the honest response is that the program was not the right fit this year, not that the child failed: identification rates run around the top few percent by design, and most bright children are not selected.
Most districts test in one of two windows: early fall (October and November), so results can shape spring placement decisions, or mid winter (January and February) for programs starting the following school year. Universal screening, where every child in a grade is tested rather than only nominated children, most often happens in Grade 2 or Grade 3, which is exactly the window this practice test targets. Districts that rely on nomination test smaller groups at more varied times.
The test is group-administered in the classroom or a computer lab, in two or three sessions on different days, with each battery taking roughly 30 to 45 minutes at full length. Teachers proctor from a script. Children who qualify for accommodations (extended time, small-group settings) generally receive them, and parents of multilingual learners should know that some districts weight the Nonverbal battery more heavily, or use the NNAT instead, precisely because it does not depend on English.
Results typically arrive four to eight weeks later as a printed or emailed report showing SAS, percentile, and stanine per battery. If the report arrives without explanation, the score section above decodes it, and your district's gifted coordinator can tell you what happens next in the identification process. Parents can usually request the scores even when no program decision follows, and they are worth keeping: a battery profile is useful context for any teacher your child meets.
Written by Michael Hodge · Founder & Assessment Director
Bachelor of Science (Psychology), University of Wollongong · LinkedIn
Last updated:
Not formally, though they are close cousins. The CogAT measures the same broad reasoning abilities that IQ tests measure, and scores on the two correlate strongly. The difference is purpose and norming: IQ tests are individually administered clinical instruments, while the CogAT is a group test normed for school-age children and built for school decisions. A CogAT SAS is not an IQ score, even though both scales center on 100.
It depends entirely on the district. Selective full-time programs often screen around the 98th percentile (SAS near 132), broader enrichment programs commonly use the 90th to 95th, and some districts qualify children on a single strong battery rather than the composite. Ask your district for its written identification policy; that document, not any website, is the real answer.
Format familiarity is not cheating, and the test's own design assumes children understand the question types. A child who has seen a figure matrix before is being measured on reasoning; a child who has not is partly being measured on decoding instructions. What crosses the line is memorizing real test items or drilling so intensively that the score stops reflecting the child. A few relaxed practice sessions sit comfortably on the right side.
Official administrations are controlled by the school, and most districts test once a year at most, with some offering a retest or appeal window. There is no personal retake option like an adult certification exam. This practice simulation, on the other hand, can be retaken freely; attempts are saved on your device so you can see scores move as formats become familiar.
That is common and genuinely informative. An uneven profile (say, a very high Quantitative score with average Verbal) describes how your child currently reasons, and many districts act on single-battery strength for subject placement. It is also a picture of the moment, not a ceiling: verbal reasoning in particular moves with reading exposure. The per-battery bars on our report exist precisely to show this profile.
The full 54-question simulation, the estimated score report, and every worked explanation are free, with no signup and no email. The site earns money from ads on the results page and from an optional paid practice pack, both clearly visible when you meet them. Nothing about your child's performance is collected or stored on our servers.
It is written for the Grade 2 to 4 screening window, roughly ages 7 to 9, with stems a young reader can manage alone. Younger children can take it with a parent reading each question aloud, which mirrors how the real primary-level test works, though the score should be read loosely. For children in Grade 5 and up the questions will feel easy, and the value is format practice rather than the number.
In school, by teachers, in group sessions, either on computers or paper. The full form runs roughly 30 to 45 minutes per battery, usually split across two or three days. At Levels 5/6 through 8 the questions are read aloud and picture-based. Our simulation deliberately relaxes the format for home practice: it is untimed, on one device, and designed to be sat together.
Treat it as a well-informed estimate, not a measurement. The percentiles come from an authored difficulty model rather than a standardization sample of real children, which is why every number on the report is labeled estimated. It is accurate enough to show relative strengths across the three batteries and to track improvement between attempts, and honest enough to say an official score can land meaningfully higher or lower.
No. The test is scored in your browser, the results live in the URL of the report page, and retake history is kept only in your own device's storage. There is no account, no email capture, and nothing identifying your child ever reaches our servers.
33 ICAR-based questions with an IQ score, percentile, and cognitive profile (for teens and adults).
40 questions across five reasoning domains with a full scored report and worked answers.
50 questions against a hard 12-minute clock, scored like the real exam.
10 sample questions with worked answers to warm up on.
For hiring teams
Send research-backed assessments to candidates by email, score them against your role profile, and review fit in one dashboard. Free to start, no credit card.
A free scored simulation of all three batteries, a worked answer for every one of the 54 questions, and no signup, email, or paywall anywhere.
Take CogAT TestThis is a free CogAT practice test with estimated, practice-level scores, built for familiarity and confidence. It is not the official CogAT and contains no real test items. CogAT is a trademark of Riverside Assessments, LLC (Riverside Insights), which is not affiliated with and does not endorse this site.