Gray's reinforcement sensitivity theory (RST) is one of the most influential contemporary theories of personality applied to the study of psychopathology. However, recent revisions to the theory have implications for known associations between personality vulnerabilities and psychopathology when using measures based on the original theory. A proposed re-structuring of a commonly used measure of the original RST, the BIS/BAS scales, as a proxy measure of the revised theory has met with some support in English-speaking countries. We sought to test the robustness of the new structure in a large non-English-speaking community sample (N = 968) and a heroin-dependent sample (N = 302). Results of exploratory and confirmatory factor analysis supported the proposed 5-factor solution comprising of two "BIS" factors (BIS-Anx and BIS-Fear) and three BAS factors (Drive, Reward responsiveness, Fun seeking) as the best fitting model for both community and heroin-dependent. However, this model showed considerable ill-fit in both samples. These findings support calls for the development of new measures consistent with the revised RST for use in applied research.