We examined the validity of the Addiction Severity Index (ASI) regarding the identification of lifetime physical and sexual abuse histories using the Structured Trauma Interview (STI) as external criterion in alcohol-dependent patients (n=144). Compared to the STI, the ASI showed a lower incidence of lifetime physical abuse reports (51% vs. 24%) and lifetime sexual abuse reports (29% vs. 17%). Lower incidence of abuse reports was stronger in males compared to females, which could be largely explained by ASI perpetrator restrictions (i.e. exclusion of several extrafamilial perpetrators). Controlling for these restrictions, acceptable sensitivity for both sexual and physical abuse as well as good specificity was found. Data indicated no response bias on the ASI in terms of social desirability or abuse severity. Sensitivity of the ASI method can probably be improved by including an opening preface to the subsequent abuse questions, including questions inquiring about abuse histories that have neutral wording instead of using the word "abuse," and inclusion of all possible perpetrators.