We studied 371 psychiatric admissions in an attempt to relate primary and secondary diagnosis, psychopathology, and stated reasons for drinking to alcohol consumption and alcoholism. No diagnostic group other than the alcoholics drank significantly more than the mean, and the schizophrenics drank less. One-third of those with bipolar and minor depression increased their drinking in the month before admission, but this was almost offset by those with similar diagnoses who drank less. Alcoholism, though not alcohol consumption, was related to a high prevalence of affective syndromes, but these did not amount to secondary diagnoses. Drinking for ‘escape reasons' was significantly associated with consumption levels, but the reasons patients gave did not correspond well with the objective presence or severity of the associated psychiatric symptoms.