Some authors advocate a broadening of the narrow concept of mixed episodes in the direction of mania leading to the concept of mixed mania, and in the direction of depression leading to the concept of mixed depression. The latter has been little investigated so far. In the present article, we retrospectively compare 49 patients with pure depression with 51 patients with mixed depression in terms of socio-demographic and clinical variables in order to contribute to the validation of the distinction between mixed and pure depression. Supporting this distinction, we observed that mixed depressive patients more frequently had past histories of bipolar disorder and alcohol abuse and had longer durations of hospital stay. These last two points remain significant even when we control for the effect of the association with bipolarity.