Most US gifted programs use an IQ score of 130 — the 98th percentile — as the standard identification threshold, combined with at least one other criterion such as achievement scores, teacher checklists, or portfolio review. Card and Giuliano (2016) demonstrated in a large Florida district that shifting from parent-and-teacher referrals to universal screening produced a 130% rise in Hispanic gifted identification and an 80% rise in Black identification, with no change to the underlying IQ standard. That's a study you should know if you care about how the current system works, because it names the mechanism: the tests aren't the whole story. Who gets tested is.
Gifted Programs and IQ Testing — Key Numbers
To see where your own reasoning profile sits against population norms across multiple domains, the DesperateMinds Advanced IQ Test scores open-answer reasoning with AI, not just multiple-choice items — a format that captures what most group-administered gifted screeners can't.
How Gifted Identification Actually Works
A single number rarely gets a child into a gifted program anymore. In most US districts today, identification runs on a two-step pipeline: an initial screen (a group-administered ability test, teacher nomination form, or achievement percentile) that flags a large pool of candidates, and then an individually administered cognitive assessment — usually the WISC-V or the Stanford-Binet 5 — that produces the score used for the eligibility decision. Both steps involve human judgment somewhere. That's where most of the trouble starts.
The 2020 State of the States in Gifted Education report noted that no US state relies on a single IQ score alone to determine eligibility. Approximately 54% of states use a multi-criteria model, in which a student must clear thresholds on two or more independent measures. Another 32% of states set no specific tests or cutoffs at all, leaving the choice to districts. The remaining states use single-cutoff-with-flexibility approaches, allowing districts to choose which test and which threshold count as sufficient.
What that means in practice is that "gifted" is a bureaucratic category before it's a psychometric one. The identification process is designed to be robust to the noise inherent in any one measurement, but it also carries whatever biases live in the measurements being combined. Adding a teacher recommendation form to an IQ score doesn't necessarily correct for bias — it can compound it.
What IQ Cutoff Does a Gifted Program Use?
The default answer is 130. That's the 98th percentile on any Wechsler-family IQ test, and it lines up with the traditional definition of "moderately gifted." Some districts hold to that number strictly. Many now accept lower thresholds — 125 (95th percentile), 120 (91st percentile), and occasionally 115 (84th percentile) — either as general policy or as a "Plan B" pathway for students from underrepresented groups.
Pennsylvania's gifted-education code is representative of the current mainstream. A student with an IQ of 130 or above qualifies automatically, but a student with a lower IQ may still be admitted if multiple criteria in their profile indicate gifted ability. Just as important, the code explicitly bars districts from denying eligibility on the basis of low working memory or processing-speed subtest scores alone. That's the WISC-V's working memory index problem in policy form — a recognition that a bright child with an executive-function weakness can end up with a lower Full Scale IQ than their reasoning warrants.
| Cutoff | Percentile | Typical Use |
|---|---|---|
| 130 | 98th | Standard automatic-eligibility threshold |
| 125 | 95th | Multi-criteria districts, some state minimums |
| 120 | 91st | Enrichment/talent-pool thresholds |
| 115 – 116 | 84th – 86th | "Plan B" thresholds for ELL / FRL students |
| 140+ | 99.6th | Highly gifted programs, Mensa eligibility |
Mensa's threshold sits above the mainstream gifted-program bar for a different reason — it's a membership society, not an educational placement, and it uses the 98th percentile on approved tests rather than a fixed IQ number. For the specifics of that separate track, the DesperateMinds guide on Mensa IQ requirements is the shortest way to see how the two systems differ. Gifted-program cutoffs, by contrast, are set by state education agencies and vary by district.
One thing worth being direct about: 130 is not a hard biological boundary. It's a convention. The bell curve is smooth, not stepped, and a child at 128 differs from a child at 132 by exactly nothing you could detect from watching them think. The cutoff exists to allocate a scarce educational resource, not to describe a difference in kind.
Which Tests Are Used for Gifted Identification?
The tests fall into two categories. Group-administered ability tests, taken by whole classrooms in a single sitting, do the initial screening. Individually administered tests, given by a licensed school psychologist, produce the score that goes into the eligibility file.
On the group side, the CogAT (Cognitive Abilities Test) is the most widely used gifted screener in the United States, delivering three separate battery scores: verbal, quantitative, and non-verbal. The NNAT (Naglieri Nonverbal Ability Test) is the second-most common, favoured for districts with large English-learner populations because it uses only visual patterns. The Otis-Lennon School Ability Test and the InView remain in use in some regions. Raven's Progressive Matrices is often cited but less commonly used as a full district screener.
On the individual side, the WISC-V (Wechsler Intelligence Scale for Children, Fifth Edition) is the field standard, with the Stanford-Binet Intelligence Scales, Fifth Edition (SB-5) in second place. Both produce Full Scale IQ scores plus multiple index scores. The KBIT-2 is used as a brief individual screener in districts that cannot afford a full WISC-V administration for every candidate. For a fuller breakdown of how these instruments compute their scores and what the indices actually measure, the DesperateMinds article on how IQ tests are scored covers the mechanics.
The choice of test matters more than most parents realise. NNAT scores and WISC-V Verbal Comprehension scores don't measure the same construct; a child who scores in the 99th percentile on one can fall in the 80s on the other. Districts that use both a non-verbal screener and a verbal-heavy individual test are, in effect, running a two-part gate that lets in students strong on either — a design that increases the pool of identified students without lowering standards on either measure.
Why Referral Systems Miss So Many Kids
Here is the counterintuitive finding the field has been slow to metabolise. Adding more tests, or making the tests fairer, doesn't solve the identification gap on its own — because the gap starts before the tests. Grissom and Redding (2016), using nationally representative data from the Early Childhood Longitudinal Study, showed that among elementary students with high standardised test scores, Black students were about half as likely as White students to be assigned to gifted services. Nicholson-Crotty, Grissom, and colleagues (2016) then documented that Black students with comparable test scores were only about a third as likely as White peers to be placed in gifted programmes when taught by a White teacher — and roughly as likely as White peers when taught by a Black teacher.
If IQ testing were the fair filter it's supposed to be, why would Black students with top standardised scores still end up half as likely to be flagged as gifted? The answer runs through the referral step. In a referral-based system, a child gets tested only if a parent or teacher nominates them. The tests themselves may or may not be biased, but the decision about who reaches the testing room is entirely discretionary — and discretion, in aggregate, tracks the biases of the people exercising it.
Test Your Reasoning Profile Across Six Domains With AI-Evaluated Open Questions
The Advanced test scores open-answer reasoning rather than only multiple choice — the format group-administered gifted screeners can't match.
Take the Advanced Test →The gap also runs through parents. High-income families are far more likely to know that gifted programmes exist, to know how to request testing, to hire a private psychologist if they doubt the school's assessment, and to appeal negative decisions. None of that is captured by an IQ score, but all of it filters who ends up with an IQ score attached to their student record. Federal data cited in Grissom's coverage put the picture starkly: Black and Hispanic students together make up roughly 40% of the US public-school population but only about 26% of gifted-programme enrolment.
A brief historical note that most coverage skips. The modern gifted-identification apparatus is a direct institutional descendant of Lewis Terman's 1920s longitudinal study, which recruited its "gifted" sample almost entirely from White, English-speaking, middle-class California families. Terman's methods were state-of-the-art for their time and produced genuinely useful data, but the demographic composition of the sample was not incidental — it reflected who his research assistants nominated. A century later, the field is still working to undo the founding assumption that giftedness looks like the kids Terman's team could see.
The Universal Screening Alternative
Card and Giuliano's 2016 paper in PNAS is the cleanest evidence in the field. Broward County Public Schools, a large urban Florida district, introduced universal screening of all second-graders in 2005. Every child took a short non-verbal cognitive test, and any child scoring above 130 (or above 116 for students on free or reduced-price lunch or classified as English language learners) was referred for the individually administered IQ evaluation used for eligibility. The IQ threshold for gifted status itself didn't change.
What happened next mattered. Hispanic identification rose by 130%. Black identification rose by 80%. Free-lunch and ELL identification rose sharply. Girls' identification rose. And — the critical detail — the IQ scores of the newly identified students matched the IQ scores of students who had been identified under the old referral system. The screening surfaced kids who were already gifted; it did not lower the bar to admit kids who weren't.
The Card and Giuliano finding is unusually clean for a policy evaluation, but it comes from one district, and Matthew McBee's commentary (2016) on the study raised a fair qualification. The newly identified students, while comparable on IQ, had lower achievement scores on average — meaning universal screening reached bright students whose academic performance had been depressed by other factors, but didn't reach every child whose potential had been suppressed by their circumstances. The screening solved part of the problem, not all of it.
Universal screening is more expensive than referral-based systems in the short run — every student has to be tested — but it is dramatically cheaper than the alternative of continuing to lose the reasoning capacity of thousands of children per district per year. The cost argument against it has never been very strong. The political argument is what has slowed adoption.
Beyond IQ — Multi-Criteria and Talent-Development Models
Renzulli's three-ring conception of giftedness — proposed in 1978 and still influential — defines gifted behaviour as the intersection of above-average ability, task commitment, and creativity. In this model, an IQ score is one input among three, not the identification itself. Many contemporary US gifted programmes descend from Renzulli's framework: they identify a broader talent pool using multiple measures and then target specific enrichment to specific strengths.
Multi-criteria identification carries a design implication most parents don't get told. When a district uses four or five measures and requires students to clear a threshold on two or three of them, the resulting pool is more demographically diverse than any single-measure cutoff would produce — because different measures favour different subgroups. Non-verbal ability tests favour students without home English exposure. Achievement scores favour students in strong schools. Teacher checklists favour students whose behaviour maps onto teacher expectations. Combining measures gives students multiple pathways in, which is what equity in this domain actually looks like in practice. The design decision most directly relevant to fluid reasoning — the ability to solve novel problems the child hasn't seen before — is which non-verbal or matrix-reasoning subtest gets used, because that's what carries most of the fluid-g signal in the WISC-V and CogAT batteries.
"The instinct to reduce giftedness to a single number is understandable — it makes admissions decisions cleaner. But the number carries less information than the profile behind it, and every district that has moved to reading the profile has ended up with a more accurate identification of its actual gifted population."
— Adam Imran, MS Clinical Psychology · DesperateMinds
The same logic underlies why the CMIAS framework, developed by DesperateMinds founder Dr. Sarwar Naseer, reports across seven cognitive dimensions rather than a single composite — the identification decision improves when you can read the shape of the profile, not only its average. That's a design principle borrowed from clinical assessment and increasingly common in gifted evaluation as well.
What This Means for Parents Navigating the System
A few practical points that most district handbooks leave implicit. First, find out whether your district uses universal screening or a referral-based system. If it's referral-based, request testing directly. Districts are not required to volunteer this information, but they are usually required to test a child on request. Second, ask which specific tests the district uses at each stage — the CogAT, the NNAT, the WISC-V — because that determines what your child is being measured on. Third, if your child scores near but below a cutoff, ask for the general ability index (GAI) or the equivalent from a Stanford-Binet report, not only the Full Scale IQ; those scores often qualify children whose FSIQ was dragged down by working memory or processing speed.
If the district's own testing produces a borderline result, a private psychologist's evaluation is usually accepted as a second opinion. The federal gifted-identification framework is discretionary rather than statutory, which is inconvenient for advocacy but flexible for individual cases. That flexibility works for families who know how to use it. It fails everyone else.
The Bottom Line
Gifted identification is a system that inherited most of its structure from a period when nobody was asking whether the process was fair, and it has spent the last twenty years catching up. The 130 IQ cutoff remains the mainstream benchmark, but the meaningful decisions in the pipeline happen upstream of the test — in who gets referred, in which screening instrument is chosen, in whether the district screens every child or waits for a nomination. Fix the upstream steps and the downstream numbers stop looking so uneven.
Frequently Asked Questions
Most US districts use 130 (the 98th percentile) as the standard IQ cutoff for gifted eligibility. A significant minority accept 125 or use lower thresholds — often 115 to 120 — for students from disadvantaged backgrounds or English language learners. No state currently reports using a single IQ score alone as its identification criterion.
For individually administered testing, the WISC-V and the Stanford-Binet 5 are the field standards. For group screening, the CogAT (Cognitive Abilities Test) is the most widely used, followed by the NNAT for nonverbal screening, the KBIT-2 for brief individual screening, and Raven's Progressive Matrices in some districts.
In many districts, yes — particularly ones using multi-criteria identification where an IQ in the 120–129 range combined with strong achievement scores, teacher recommendations, or non-verbal reasoning strengths qualifies a student. Districts using a strict 130 cutoff on a single test would not admit a 125, but very few districts still work that way.
Grissom and Redding (2016) found Black students with high standardised test scores are about half as likely as White peers to be assigned to gifted services, and only about a third as likely when the classroom teacher is White. Referral-based identification concentrates discretion in individuals whose implicit judgments reproduce existing inequities.
Universal screening means every student in a grade is tested — not only those referred by parents or teachers. Card and Giuliano (2016) documented that shifting to universal screening in a large Florida district increased Hispanic gifted identification by 130% and Black identification by 80%, with no lowering of the underlying IQ threshold.
In many districts, yes. National surveys show 32% of states set no specific test or cutoff at all, and most others allow alternative pathways — group ability tests, achievement scores, portfolio review, or teacher checklists — instead of an individually administered IQ measure. Even where an IQ test is required, districts often accept multiple test brands.
Well-administered individual IQ tests are highly reliable at the population level but noisier at the individual level, particularly near cut-offs. Standard error of measurement means a child's true score sits inside a roughly 4-to-6-point band around the obtained score. Multi-criteria identification is a direct response to this measurement limitation.
See Your Verbal, Numerical, and Spatial Reasoning Under Adaptive AI Scoring
The Advanced Test evaluates open-response items across six cognitive domains — a fuller read of the reasoning profile than a group-administered screener produces.
Start the Advanced Test →References
- Card, D., & Giuliano, L. (2016). Universal screening increases the representation of low-income and minority students in gifted education. Proceedings of the National Academy of Sciences, 113(48), 13678–13683.
- Card, D., & Giuliano, L. (2015). Can universal screening increase the representation of low income and minority students in gifted education? NBER Working Paper No. 21519.
- Grissom, J. A., & Redding, C. (2016). Discretion and disproportionality: Explaining the underrepresentation of high-achieving students of color in gifted programs. AERA Open, 2(1), 1–25.
- Nicholson-Crotty, S., Grissom, J. A., Nicholson-Crotty, J., & Redding, C. (2016). Disentangling the causal mechanisms of representative bureaucracy: Evidence from assignment of students to gifted programs. Journal of Public Administration Research and Theory.
- Lakin, J. M. (2016). Universal screening and the representation of historically underrepresented minority students in gifted education. Journal of Advanced Academics, 27(2), 139–149.
- Renzulli, J. S. (1978). What makes giftedness? Reexamining a definition. Phi Delta Kappan, 60, 180–184.
- NAGC. (2018). Use of the WISC-V for Gifted and Twice Exceptional Identification. National Association for Gifted Children Position Statement.
- McBee, M. T. (2016). What you don't look for, you won't find: A commentary on Card and Giuliano's examination of universal screening. Journal of Advanced Academics, 27(3), 218–222.