Across 6,507 people who have taken the Dark Mirror quiz, 298 of them, 4.6%, lead on the narcissistic axis, and only 2 of the 6,507 score 60 or above on it. That second figure, two people in six and a half thousand, is the most interesting number on this page, and one honest reading of it is that our own instrument may be the reason.
What follows is our data first, clearly labelled, then the published clinical research, kept separate from it. The two measure different things and must never be added together or read as rival estimates of the same quantity.
Our figures: the narcissistic axis on the Dark Mirror
Edition computed 5 September 2026, sample 6,507 takers.
- 298 takers (4.6%) had narcissistic as their leading axis.
- 2 takers (0.03%) scored 60 or above on the narcissistic axis.
- Mean score 10.8 out of 100, which is 2.16 keyed answers out of 20. Standard deviation 9.5 points.
- Internal consistency KR-20 of 0.51, the second weakest of the six axes. Only histrionic, at 0.42, is lower.
- Among the 298 narcissistic-leading takers, the most common second axis was psychopathic (109), then sociopathic (107), then neurotypical (52), borderline (20) and histrionic (10).
Correlations across all 6,507 takers, narcissistic against each of the other axes:
- narcissistic and sociopathic, r = 0.26
- narcissistic and histrionic, r = 0.24
- narcissistic and psychopathic, r = 0.18
- narcissistic and borderline, r = -0.12
- narcissistic and neurotypical, r = -0.60
The only strong relationship there is the negative one with the neurotypical axis, which is close to arithmetic: keyed answers that go to one axis cannot go to another. The mild positive links to sociopathic and histrionic are the sort of overlap the personality literature would expect, and they are mild. The small negative link with borderline is worth noting mainly because so many popular accounts treat the two as neighbours.
Where narcissism sits among the six axes
Leading axis for the same 6,507 takers:
| Leading axis | Takers | Share of takers |
|---|---|---|
| Neurotypical | 3,816 | 58.6% |
| Psychopathic | 1,365 | 21.0% |
| Sociopathic | 594 | 9.1% |
| Borderline | 419 | 6.4% |
| Narcissistic | 298 | 4.6% |
| Histrionic | 15 | 0.2% |
Every one of those percentages is a share of people who took this quiz. None is a rate in the general population, and the sample is not a community sample. More on that below.
Almost nobody maxes narcissism, and we do not know why
Two takers out of 6,507 reached the high band on the narcissistic axis. On an audience that arrives through dark psychology content, and therefore skews heavily towards people already interested in these traits, that is a striking result. It deserves better than a shrug.
There are three candidate explanations. We cannot tell you which is right, and the third one is about us.
One: the items are transparent. Self-report narcissism questions announce what they measure. A reader can see, in the moment of answering, what endorsing an item says about them, and grandiosity is socially undesirable to admit even to an anonymous form. Nothing about anonymity removes the audience of one, which is the person answering.
Two: the people who would score high do not arrive. Scanning yourself for a trait requires suspecting you have it. The disposition most associated with a high narcissistic score is not the disposition most likely to seek out a quiz about narcissism. If that is the mechanism, the ceiling is a selection artefact and says nothing about how the trait is distributed anywhere else.
Three: our instrument may simply be weak on this axis. KR-20 of 0.51 is modest. It means the items do not hang together very tightly when they measure narcissism, and a set of items that do not agree with one another will rarely stack twelve deep in one person. A low ceiling can be a property of the questions rather than of the people answering them. We report the figure rather than bury it, and we would rather this page said so than had the finding read back to us later.
Readers who want the shape of the whole distribution rather than the tail can find it in the average narcissist score.
Published research: what the clinical literature says
These are other people, other methods, other samples. They are not our numbers, and they are not comparable to ours.
- Narcissistic personality disorder: 6.2% lifetime prevalence. Stinson et al. 2008, Journal of Clinical Psychiatry, 34,653 US adults, second wave of the National Epidemiologic Survey on Alcohol and Related Conditions, structured diagnostic interview. Point prevalence is usually quoted nearer 1%. PubMed 18557663
- Antisocial personality disorder: 3.6% lifetime prevalence. Grant et al. 2004, 43,093 US adults, first wave of the same survey programme. PubMed 15291684
- Borderline personality disorder: 5.9% lifetime prevalence. Grant et al. 2008, second wave. PubMed 18426259
A clinical diagnosis in that literature turns on pervasiveness across situations, duration over years, and impairment or distress. A quiz axis turns on how many of twenty scenarios a person answered one particular way. Those are not the same object measured at different precision. They are different objects.
The inversion, stated as a question
Set the two blocks side by side and something odd appears.
In the diagnostic literature, narcissistic personality disorder is the more common lifetime diagnosis of the two, 6.2% against 3.6% for antisocial personality disorder. On our quiz, the narcissistic axis is the one people are least likely to max, while the psychopathic axis leads for 21.0% of takers and the sociopathic axis for another 9.1%.
We are not claiming that this proves anything about either measure. It is a question the data raises, and there are at least four ways it could resolve: our audience is unusual, our narcissism items are weak, self-report suppresses this trait harder than it suppresses coldness, or the constructs do not line up between a screener and a clinical interview. Possibly all four at once. Anyone who tells you which one it is, on this evidence, is guessing.
If the distinction between those two shapes is what brought you here, narcissist versus sociopath sets out the behavioural differences without the arithmetic.
How to read any of these numbers
Four rules, and they are the reason this page exists in the form it does.
Our figures describe a self-selected audience. People arrive at the Dark Mirror from a dark psychology creator site, mostly through social platforms and search. That is not a community sample, and no share we publish is a prevalence estimate. Every one of our figures means "of people who took this quiz".
A screener is not an interview. Twenty scenarios answered in a few minutes cannot do what a trained clinician does across an hour with corroborating history.
Bands are a reading aid. Baseline 0 to 15, traits 20 to 35, pronounced 40 to 55, high 60 and above, which is twelve of twenty answers. Those cut points exist so that a number has a sentence attached to it. They are not clinical thresholds, and no band is a diagnosis of anything.
Suspicion about yourself is not evidence about yourself, in either direction. People who worry that they are narcissistic often are not, and a low score is not a clearance certificate. If the question is personal rather than statistical, the signs you might be a narcissist and the difference between covert and overt presentations are more use than a percentile.
Method, in short
The Dark Mirror is 20 questions. Each answer is keyed to one of six axes, and an axis score is the count of answers keyed to it multiplied by five, giving 0 to 100 in steps of five. The six axes are psychopathic, sociopathic, narcissistic, borderline, histrionic and neurotypical.
For this edition, 6,526 takes were recorded and 6,507 takers form the published sample. Four same-account retakes and 15 takes completed in under 60 seconds were excluded. A further 224 straight-line answer sets are flagged in the published exclusion ledger. Internal consistency is reported per axis as KR-20; narcissistic is 0.51.
The full instrument description, exclusion ledger, reliability figures and confidence intervals are on the methodology page, and the live distributions are on the Dark Mirror Index.
The figures, numbered for citation
- Among 6,507 Dark Mirror takers, 298 (4.6%) had the narcissistic axis as their leading axis.
- Two takers (0.03%) scored 60 or above on the narcissistic axis, the high band, which requires 12 of 20 keyed answers.
- The mean narcissistic score is 10.8 out of 100, with a standard deviation of 9.5 points.
- Internal consistency on the narcissistic axis is KR-20 0.51, the second lowest of the six axes; histrionic is lowest at 0.42.
- Among narcissistic-leading takers, the most common second axis was psychopathic (109 of 298), then sociopathic (107 of 298).
- The narcissistic axis correlates at r = 0.26 with sociopathic, r = 0.24 with histrionic, r = 0.18 with psychopathic, r = -0.12 with borderline and r = -0.60 with neurotypical.
- Leading axis across the sample: neurotypical 58.6%, psychopathic 21.0%, sociopathic 9.1%, borderline 6.4%, narcissistic 4.6%, histrionic 0.2%.
- All seven figures above describe people who took this quiz. None is a population prevalence estimate.
- For comparison from the clinical literature, lifetime prevalence of narcissistic personality disorder is 6.2% (Stinson et al. 2008, n = 34,653), against 3.6% for antisocial personality disorder (Grant et al. 2004) and 5.9% for borderline personality disorder (Grant et al. 2008).
Figures dated 5 September 2026. The Index recomputes daily, so the shares move as the sample grows.
Frequently Asked Questions
How many people score high on narcissism?
On the Dark Mirror, 2 takers out of 6,507 scored 60 or above on the narcissistic axis, which is 0.03% of people who took this quiz. Another 298, or 4.6%, led on that axis without reaching the high band. These are figures for a self-selected audience, not a population rate, and no band is a diagnosis.
How common is narcissistic personality disorder?
The largest US epidemiological estimate, Stinson and colleagues in 2008, put lifetime prevalence of narcissistic personality disorder at 6.2% among 34,653 adults interviewed in the second wave of NESARC. Point prevalence is usually quoted nearer 1%. Those figures come from structured diagnostic interviews in a community sample, which is a different measurement from any online screener, including this one.
What is the average narcissism score on the Dark Mirror?
The mean narcissistic score is 10.8 out of 100, with a standard deviation of 9.5 points. Scores move in steps of five, so 10.8 is about 2.16 keyed answers out of 20. Most takers sit in the baseline band of 0 to 15. A score of 40 or more is uncommon in this sample, and 60 or more is close to absent.
Is a narcissism quiz result a diagnosis?
No. The Dark Mirror is a 20 question self-report screener, and its four bands, baseline, traits, pronounced and high, are a reading aid rather than clinical thresholds. Narcissistic personality disorder is diagnosed by a clinician, usually after a structured interview covering pervasiveness, duration and impairment. A quiz score describes the answers somebody gave on one day, nothing more.