This study relates to the validity of the Self-Report Psychopathy Scale (SRP), in French-speaking Belgium, in community, forensic and clinical samples. Against this background, we set out to assess its discriminant validity on Belgian participants from three settings: community (Non-Offender/Non-Psychiatric; N=18), general psychiatry (Non-Offender/Psychiatric; N=23), and forensic psychiatry (Offender/Psychiatric; N=25). We also evaluated, on all participants, the SRP-III’s convergent validity with dimensions considered characteristic of psychopathy under the Hare (2003) model. The results show the SRP-III to be more in line with the literature in terms of behavioral dimensions, such has impulsivity and aggression, than of personality traits. They also reveal that the SRP-III is better at discriminating psychiatric populations than non-psychiatric populations.