The factor structure of the Beck Depression Inventory (BDI) is still controversial. The present study attempted to replicate a general two-factor model (cognitive-affective and somatic factors) of the BDI with a confirmatory Procrustes rotation procedure in two clinical samples of patients being treated for alcohol dependence (N₁ = 243, N₂ = 148) and one clinical sample of patients admitted to an acute psychiatric ward because of an act of deliberate self-harm (N₃ = 144). In addition, due to the neglect of gender-specific analysis in factor-analytic studies of the BDI, the two-factor model was tested for males and females separately. The results did not support the validity of the two-factor model in the total samples. Gender-specific findings indicated a better fit of the model in male samples. Possible implications of a gender-specific factor structure of the BDI are discussed.