Background. Gender differences in the prevalence of post-traumatic stress disorder were examined by analysing discrepancies between the DSM-IV and ICD-10 diagnostic systems. Method. Data from the Australian National Survey of Mental Health and Well-Being (n=10641) were analysed at the diagnostic, criterion and symptom level for DSM-IV and ICD-10 PTSD for males versus Females. Results. While there was a significant gender difference in the prevalence of PTSD for ICD-10, no such difference was found for DSM-IV. The pattern of gender difference at the diagnostic level was mirrored in the pattern of gender differences at the criterion level for both DSM-IV and ICD-10. Females only endorsed three symptoms at a significantly higher rate than males. For all other symptoms, endorsement was equal. This apparently small gender difference at the symptom level was sufficient to cause the gender difference at the diagnostic level for ICD-10, but not DSM-IV because of the different manner in which symptoms are configured into criteria in each of the diagnostic systems. Conclusions. Gender differences in ICD-10 PTSD but not in DSM PTSD diagnoses are attributable in this study to different patterns of endorsement of symptoms by males and females. Possible reasons for the differential endorsement of symptoms and implications for the use of epidemiological instruments are discussed.