Nearly every personality-type percentage on the internet traces back to one national standardization sample, published in 1998. In the 2025 synthesis that reproduces it, three printed cells fail their own arithmetic. This page carries the corrected board — by type, and by sex.
Tick length tracks how common each type is overall. † marks a figure recomputed by CINEMAWORDS.
Select any type for how rare it is, how its cell splits between men and women, and its five pieces on this site.
Every serious count of the sixteen types across the American adult population comes from one source: a national standardization sample of 3,009 adults, first published in 1998. Of those, 1,478 were men and 1,531 women. No comparable standardization sample has been published since. Everything newer is aggregation. A 2025 synthesis pooled the samples published between 1999 and 2024, the twenty-five years that followed. Across more than forty-four thousand people counted by full four-letter type, it found proportions leaning consistently toward Intuition and Thinking. Study samples are volunteers, students, and clients, not a cross-section of anybody.
There is one more problem, and it is the reason this page’s numbers differ from most copies of the same table. In the 2025 synthesis’s reproduction of the women’s column, three cells disagree with the frequency counts printed beside them. Thirteen INTJ women out of 1,531 is 0.8 percent, not the 0.9 that appears in print. The ISTP cell fails the same check upward, 2.4 rather than 2.3, and so does INTP, 1.8 rather than 1.7.
The counts are the sounder column. Add any type’s two counts, divide by 3,009, and you land on its overall percentage, sixteen times out of sixteen. The columns themselves sum to exactly 1,478 men and 1,531 women. The printed women’s column sums to 99.9. Correcting these three cells brings it to exactly 100; the men’s column already lands there. So where a printed percentage and its own frequency disagree, this page carries the frequency. The three corrected cells are marked in the full table below.
What is the rarest personality type?
Overall, INFJ — one in sixty-seven adults. But “rarest” splits the moment you split the sample. Among men the rarest type is still INFJ, at 1.3 percent. Among women it is INTJ, at 0.8 percent — thirteen women in the sample of 1,531, the smallest cell anywhere on the board. A woman who tests INTJ is rarer in this sample than an INFJ of either sex.
What is the most common personality type?
Overall, ISFJ — roughly one adult in seven. Split the sample and the leaders diverge: among men it is ISTJ, at one in six; among women it is ISFJ, at nearly one in five. One caution about newer numbers. In the twenty-five years of self-selected study samples, the most common type among women is ENFP, at 13.7 percent. That is a fact about who volunteers for studies, not about the population, and this page keeps the two datasets apart.
Which types differ most between men and women?
Rank the board by tilt and the male-heavy end runs INTJ, ISTP, ENTJ, INTP. INTJ is the steepest ratio anywhere — close to four men to one woman. ISTP follows at just over three and a half, ENTJ at three, INTP at two and three-quarters. On the printed percentages the first two swap places; on the frequency counts they do not, and this page follows the counts. The female-heavy end mirrors it: ISFJ at nearly two and a half women to one man, ESFJ just behind, ENFJ at two to one. At the other extreme the columns nearly touch. INFP is the flattest cell on the board, with INFJ and ISFP close behind. All four of the biggest gaps are Thinking types; the three steepest reversed gaps are all Feeling–Judging — the tilt follows the T/F letter far more than any other.
Why do these numbers disagree with what I see online?
Three reasons, in descending size. First, most online statistics quote self-selected test-takers, and people who seek out personality tests skew measurably toward Intuition and away from Sensing — the 2025 aggregate shows the same drift. Second, many sites copy each other’s copies, including the three misprinted cells this page corrects. Third, the categories themselves are soft at the joints. In the four-week retest reported in the instrument’s own manual, only about two people in three reproduced the same four letters. An individual’s cell membership wobbles even before sampling does.
How reliable are these percentages?
Reliable for what they are; narrow in what they cover. The sample is American, English-language, and from 1998 — a landscape, not a prophecy, and certainly not a verdict on any individual. The four underlying scales measure consistently (four-week score reliability runs .85 to .91), but the step from four scores to sixteen boxes is where the softness lives. Read the table as the honest base rate for a question like “how unusual is this type, for this sex” — and read nothing in it as a statement about what any one person can do.
Does being a rare type mean anything?
Not in the way the question usually means it. A rarity figure says how often a four-letter combination turned up in one sample of three thousand Americans, and that is the whole of what it says. It carries nothing about capability, fit, or worth, and it does not get more meaningful the smaller the cell gets. There are about twelve ISFP women in that sample for every INTJ woman. She is not, on that account, twelve times anything else. The number describes a crowd. It was never about a person.
There is also the wobble underneath. About a third of people who sit the questionnaire twice, four weeks apart, come back with at least one letter moved. Some share of every cell on this board would land elsewhere on a second sitting. Rarity is a property of the tally, not of the reader.
The full table: all sixteen types, by sex
| Type | Overall % | Men % (n) | Women % (n) | Tilt (M:W) |
|---|---|---|---|---|
| ISTJ | 11.6 | 16.4 (242) | 6.9 (106) | 2.4 : 1 |
| ISFJ | 13.8 | 8.1 (119) | 19.4 (297) | 1 : 2.4 |
| INFJ | 1.5 | 1.3 (19) | 1.6 (25) | 1 : 1.3 |
| INTJ | 2.1 | 3.3 (49) | 0.8† (13) | 3.9 : 1 |
| ISTP | 5.4 | 8.5 (126) | 2.4† (36) | 3.6 : 1 |
| ISFP | 8.8 | 7.6 (112) | 9.9 (152) | 1 : 1.3 |
| INFP | 4.4 | 4.1 (61) | 4.6 (71) | 1 : 1.1 |
| INTP | 3.3 | 4.8 (71) | 1.8† (27) | 2.7 : 1 |
| ESTP | 4.3 | 5.6 (83) | 3.0 (46) | 1.9 : 1 |
| ESFP | 8.5 | 6.9 (102) | 10.1 (154) | 1 : 1.5 |
| ENFP | 8.1 | 6.4 (95) | 9.7 (148) | 1 : 1.5 |
| ENTP | 3.2 | 4.0 (59) | 2.4 (37) | 1.7 : 1 |
| ESTJ | 8.7 | 11.2 (165) | 6.3 (96) | 1.8 : 1 |
| ESFJ | 12.3 | 7.5 (111) | 16.9 (259) | 1 : 2.3 |
| ENFJ | 2.5 | 1.6 (24) | 3.3 (50) | 1 : 2.0 |
| ENTJ | 1.8 | 2.7 (40) | 0.9 (14) | 3.0 : 1 |
† The published percentage for this cell disagrees with the frequency count printed beside it in the source; the figure shown is recomputed from the count. Percentages are counts over 3,009 adults, 1,478 men, 1,531 women. Tilt is a type’s rate among men divided by its rate among women, computed from the counts rather than from the rounded percentages beside it. Source: the 1998 national standardization sample (n = 3,009) as reproduced in Bradley T. Erford et al., Journal of Counseling & Development 103 (2025): 403–417, doi:10.1002/jcad.70006. © 2025 The Author(s); published by Wiley Periodicals LLC on behalf of the American Counseling Association. Table compiled by CINEMAWORDS; source data reused and adapted under CC BY 4.0, the three marked cells being the adaptation. The twenty-five-year study aggregate (n = 44,278) referenced above is reported separately in the same synthesis and is self-selected.
Every type, in full
Each type on this board has a full profile and four companion pieces.