How Reliable Is the Big Five Personality Test?
Take a personality test today and again next month — should you expect the same result? For the Big Five, the answer is largely yes: peer-reviewed studies put test-retest reliability between .75 and .90 on standard instruments.[1] This page explains what those numbers mean, why MBTI scores so much lower, and what genuinely affects your results.
The quick answer
Big Five test-retest reliability is high. On established instruments — IPIP-50, NEO PI-R, BFI-2 — your scores correlate .75 to .90 across intervals of weeks to months.[1][2][3]
MBTI test-retest reliability is poor. Studies have found that 35–50% of people receive a different four-letter type when retaking the MBTI within five weeks.[4]
The difference is structural, not random. Big Five uses continuous scales; MBTI uses binary categories. Continuous scores absorb small fluctuations gracefully. Binary categories flip dramatically when someone scores near the midpoint.
What "test-retest reliability" actually means
Test-retest reliability measures whether you get a similar result when you retake the same test under similar conditions. It is expressed as a correlation coefficient between two testing occasions, usually written as r:
- •r = 1.0 — perfectly identical scores both times.
- •r = 0.0 — scores are completely unrelated; the test is essentially random.
- •r ≥ 0.70 — the conventional threshold psychometricians treat as "good enough" for a self-report measure.
Two things to keep in mind. First, reliability is not the same as validity — a test can be highly reliable (consistent) without measuring what it claims to measure. Second, reliability is interval-dependent: a test measured days apart will look more reliable than the same test measured years apart, simply because real personality change accumulates over time.[5]
The reliability numbers you will see in the rest of this article are typically based on intervals of weeks to months — short enough that genuine personality change is minimal, but long enough that you are not just remembering your previous answers.
The numbers: Big Five scores by instrument
Different Big Five instruments report slightly different reliability values depending on length, scoring method, and study design. Here are the typical ranges from peer-reviewed research:
| Instrument | Items | Test-retest r | Source |
|---|---|---|---|
| IPIP-50 (used by this site) | 50 | .75–.85 | Gow et al. 2005[1] |
| NEO PI-R | 240 | .85–.92 | Costa & McCrae 1992[2] |
| BFI-2 | 60 | ~.80 | Soto & John 2017[3] |
| MBTI (for comparison) | ~93 | ~.50 | Pittenger 1993[4] |
Two patterns stand out. First, longer instruments tend to be more reliable — the 240-item NEO PI-R squeezes more measurement out of each trait than the 50-item IPIP, which is why it sits higher. The trade-off is that NEO PI-R takes 30+ minutes; the IPIP-50 takes about seven.[6]
Second, every Big Five instrument outperforms MBTI by a wide margin — the gap is roughly 0.25–0.35 in correlation, which is substantial. The chart below visualizes this difference:
For our IPIP-50 in particular, you can expect that if you take the test today and again in a month, your scores on each of the five traits will be very close — typically within 5–10 percentile points unless something significant in your life has shifted.
Why MBTI reliability is so low
The MBTI's low reliability isn't primarily a measurement-error problem. It is a design choice: the MBTI converts continuous scores into binary categories at the midpoint, which inflates apparent instability.[4][7]
Imagine someone who scores 52% Extraverted on the MBTI. They get labeled E. Take the test again next week, score 49% Extraverted — now they get labeled I. Their personality didn't flip; their underlying score barely moved. But their published type changed completely, from (say) ENFP to INFP.
On the Big Five, the same person would simply move from 52% Extraversion to 49% Extraversion — a barely noticeable shift on a continuous scale, with the same overall profile.
An important nuance: the underlying continuous MBTI scores are themselves more stable than the type labels. The instability shows up only when those scores are forced through a binary cutoff. This is a known critique of forced-choice typology going back decades.[8]
For a fuller comparison of the two frameworks, see our Big Five vs MBTI guide.
Long-term stability and lifespan change
Short-term reliability is high, but personality is not static across the lifespan. This is one of the most consistent findings in personality psychology: Big Five traits change in slow, predictable patterns as people age.[5][9]
A landmark meta-analysis by Roberts, Walton & Viechtbauer (2006) tracked Big Five scores across 92 longitudinal studies and found:
- Conscientiousness rises steadily through young adulthood and middle age — most people become more responsible and organized in their 20s, 30s, and 40s.
- Agreeableness rises gradually across the lifespan, with the biggest gains in the 30s and beyond.
- Neuroticism declines through young and middle adulthood — most people become more emotionally stable as they age.
- Openness peaks in young adulthood and declines slightly after age 30, though the change is smaller than for the other traits.
- Extraversion is mostly stable, with small declines in social-vitality facets after middle age.
The implication: if you take the Big Five at age 22 and again at age 42 and your scores have shifted, that is not evidence the test is unreliable. It is evidence your personality genuinely changed — which is consistent with decades of research. The test is doing its job by reflecting that change.
What affects your retest score (besides real change)
Even within a few weeks, your score on a Big Five test isn't guaranteed to be identical. Several factors create normal, expected variation:
Your mood at the moment
Particularly visible on Neuroticism. If you take the test on a stressful day, your Neuroticism score will read a few points higher than usual. This is real signal — your emotional state genuinely affects answers — but it is short-term noise on top of your stable trait score. (More on Neuroticism →)
Recent life events
A breakup, a promotion, a move, a new diagnosis — all can temporarily shift how you describe yourself. Big Five traits are stable on average, but specific weeks can pull scores in one direction.
Social desirability bias
People sometimes answer how they wish they were rather than how they are. This affects Conscientiousness and Agreeableness most, since those traits are seen as "more virtuous." Reverse-scored items (where agreeing means a lower score) help reduce this bias, which is why most modern Big Five tests use them.
Test fatigue
On longer tests, attention drops in the second half, slightly reducing reliability. The IPIP-50 is short enough that this is rarely an issue.
Item interpretation drift
How you read a statement like "I am the life of the party" can shift between testings — what counts as "the life of the party" depends on context. This is unavoidable noise that every self-report test contains.
How to make your test result as accurate as possible
- Take it when rested and not in an emotional extreme (very anxious, very euphoric).
- Answer based on how you typically are over months and years — not how you have felt this week.
- Don't overthink. First instinct is usually closer to your real self than a deliberated answer.
- Be honest, especially on items where you might want to look better than you are. The test isn't graded — accuracy serves you, not us.
The bottom line for users
If you take a careful Big Five test like the IPIP-50, three things should be true:
- Your scores will be highly consistent within weeks. Retake in a month and expect each trait to land within a handful of percentile points of last time.
- Your scores will gradually shift over years. That is real personality change — not a test failure — and is consistent with the lifespan-development findings above.
- If you ever retake a personality test and get a completely different answer, that is much more likely to indicate the test is type-based (like MBTI) than that you are unstable.
Big Five scores aren't a horoscope or a rigid label. They are the closest thing personality psychology has to a stable, repeatable, scientifically validated self-portrait.
Frequently asked questions
If I retake the test and get different scores, which one is the real me?▾
On a Big Five test, small differences (within ~10 percentile points) are normal short-term noise — neither result is "wrong." Take the average if you want a single number. If your scores have shifted dramatically over months or years, the more recent test is probably more accurate to who you are now, since personality genuinely changes over time.
When does it make sense to retake the Big Five test?▾
Useful retake intervals are 1–2 years apart, especially around major life transitions (graduating, starting a new job, becoming a parent, going through therapy). Retaking within days of the first test mostly measures memory and answering style, not personality.
Do mood and current life events change my scores?▾
Yes — but mostly on Neuroticism. Stress and recent setbacks reliably push Neuroticism scores up by a few points. Conscientiousness and Openness are more stable across moods. The other three traits sit in the middle. If you suspect mood is biasing your test, wait a week and retake.
Why do shorter Big Five tests have lower reliability than the IPIP-50?▾
Each trait is measured by averaging across multiple items. With more items per trait, random noise on any one item gets averaged out. A 10-item Big Five test (2 items per trait) is much noisier than a 50-item one (10 items per trait). Reliability scales roughly with the square root of test length.
I took two different Big Five tests and got different scores — is that normal?▾
Different Big Five instruments use slightly different questions and reference populations, so small differences are expected. They should still produce similar overall profiles — if you score high Openness on one Big Five test, you should also score high Openness on another. If profiles are wildly different, one of the tests is likely poorly constructed.
How is reliability different from accuracy?▾
Reliability is consistency: do you get the same result twice? Accuracy (validity) is correctness: does the test actually measure what it claims? A test can be reliable without being accurate — a broken bathroom scale that always reads 5 lbs heavy is perfectly reliable but useless. The Big Five is both reliable AND accurate, supported by decades of validity research linking scores to real-world outcomes like job performance, health, and relationship quality.
Take the Big Five test now
Our free test uses the IPIP-50 — the same instrument cited in the reliability research above. 50 questions, 7 minutes, no signup, scientifically validated.
Take the Free Big Five TestKeep reading
If you found the science useful, here are the natural next stops on this site:
A side-by-side comparison of the two most popular personality frameworks.
What "OCEAN" means, where the model came from, and how each trait is measured.
Percentiles, score ranges, and what each level actually means.
Deeper guides to Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism.
References
- [1] Gow, A. J., Whiteman, M. C., Pattie, A., & Deary, I. J. (2005). Goldberg's 'IPIP' Big-Five factor markers: Internal consistency and concurrent validation in Scotland. Personality and Individual Differences, 39(2), 317–329.
- [2] Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI) Professional Manual. Psychological Assessment Resources.
- [3] Soto, C. J., & John, O. P. (2017). The next Big Five Inventory (BFI-2): Developing and assessing a hierarchical model with 15 facets to enhance bandwidth, fidelity, and predictive power. Journal of Personality and Social Psychology, 113(1), 117–143.
- [4] Pittenger, D. J. (1993). Measuring the MBTI… and coming up short. Journal of Career Planning and Employment, 54(1), 48–52.
- [5] Roberts, B. W., Walton, K. E., & Viechtbauer, W. (2006). Patterns of mean-level change in personality traits across the life course: A meta-analysis of longitudinal studies. Psychological Bulletin, 132(1), 1–25.
- [6] Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C., Cloninger, C. R., & Gough, H. G. (2006). The international personality item pool and the future of public-domain personality measures. Journal of Research in Personality, 40(1), 84–96.
- [7] Pittenger, D. J. (2005). Cautionary comments regarding the Myers-Briggs Type Indicator. Consulting Psychology Journal: Practice and Research, 57(3), 210–221.
- [8] McCrae, R. R., & Costa, P. T. (1989). Reinterpreting the Myers-Briggs Type Indicator from the perspective of the five-factor model of personality. Journal of Personality, 57(1), 17–40.
- [9] Anusic, I., & Schimmack, U. (2016). Stability and change of personality traits, self-esteem, and well-being: Introducing the meta-analytic stability and change model of retest correlations. Journal of Personality and Social Psychology, 110(5), 766–781.
- [10] Costa, P. T., Herbst, J. H., McCrae, R. R., & Siegler, I. C. (2000). Personality at midlife: Stability, intrinsic maturation, and response to life events. Assessment, 7(4), 365–378.