The aim of this research project was to study gender identification in male transsexuals compared to male and female controls, using the Rorschach test and the MMPI. In the international literature, many researches have shown that the nature of the human response on Rorschach card III is linked to gender identification, as is the MMPI Mf scale. Ten untreated male homosexual transsexuals and 18 treated and operated male homosexual transsexuals were compared to 10 male and 12 female controls regarding verbal IQ, human content on Rorschach card III and the MMPI Mf scale. Absence of hormonal treatment for the first group of transsexuals was checked by a blood test at the time of the psychological testing. Responses on Rorschach card III were scored according to different kinds of human contents: male (M), female (F), gender-unidentified/neutral (N), bisexual (B), feminine then masculine or the opposite (M/F), and nonhuman (NH). N, B, M/F and NH responses were rare in all Rorschach protocols. As expected, responses given by participants in the control group were significantly more consistent with their anatomical sex than with the opposite sex. Untreated transsexuals do not differ from treated and operated transsexuals on Rorschach data, and both transsexual groups give significantly more female human representations than male controls. Transsexuals' results are similar to female controls. Untreated transsexuals' mean score on the MMPI Mf scale is significantly higher than that of treated and operated transsexuals' score, in the male profile (biological sex). Both groups of transsexuals score higher on the Mf scale in the male profile than in the female profile. The mean Mf score in the male profile is significantly higher than that of male controls, whereas, in the female profile, the mean Mf score is similar to that of female controls. This study shows that for both groups of transsexuals, results are homogenous in respect of Rorschach and MMPI, showing hyper-conformism to self-perceived gender. Results in both groups are similar to results of female controls, but tend to show even more feminine gender identification. The absence of any significant difference between untreated and treated and operated transsexuals seems surprising, suggesting that the hormonal treatment has not had a major impact on gender identification processes. It would doubtless be interesting to study gender identification using even more kinds of data: all human contents in the Rorschach protocol (not just the responses given to card III), MMPI Mf scale, Draw-A-Person Test and Animal-and-Opposite Drawing Test. This would enhance result liability and could provide useful information about how gendter identification processes evolve after surgical sex reassigment.