In a double-blind trial, 32 patients with endogenous or neurotic depression requiring drug treatment in hospital were randomly allocated to a four-week treatment with amitriptylinoxide or amitriptyline. The initial dosage was, in most cases, 60 mg b.i.d. for both drugs. During the first week of treatment, the dose was increased to a maximum of 300 mg/day with amitriptylinoxide and to 240 mg/day with amitriptyline, depending on tolerability and clinical impression. This dose was then maintained, if possible, throughout the following three weeks. During this period the mean dose of amitriptylinoxide was 167 +/- 8 mg/day, that of amitriptyline 166 +/- 8 mg/day. Seven of the 32 patients (three in the amitriptylinoxide and four in the amitriptyline group) discontinued the trial. Therapeutic efficacy was assessed by means of the AMDP system (scale 4), HAMD total scale, 4 HAMD subscales, and DSI and Bf-S scales. Tolerability was assessed using the AMDP system (scale 5) and a list of side-effects often associated with antidepressive drug treatment. With regard to therapeutic efficacy, no differences were noted between the two drugs except on the DSI scale. In contrast, there was a significant difference in tolerability--both in terms of the vegetative syndrome (AMDP) and the sum of side-effects--in favour of amitriptylinoxide.