BACKGROUND: About one-third of patients fail to respond to initial antidepressant therapy, which suggests a need for more effective drugs. AIMS: To compare the efficacy and safety of venlafaxine and paroxetine in 122 patients with non-chronic treatment-resistant depression. METHOD: In-patients or out-patients satisfying DSM-III-R criteria for major depression in evolution for less than eight months, having a baseline HAM-D score > or = 18 and a HAM-D Item 3 score < 3 were eligible. Patients were required to have a history of resistance to two previous antidepressant treatments and a CGI improvement score of 3 at the beginning of treatment. Doses were adjusted to 200-300 mg/day for venlafaxine and 30-40 mg/day for paroxetine. RESULTS: For the observed-case analysis, the response rate was 51.9% for venlafaxine and 32.7% for paroxetine (P = 0.044), and a remission was achieved in 42.3% of venlafaxine-treated and 20.0% of paroxetine-treated patients (P = 0.01). The incidence of adverse effects was comparable between treatment groups. CONCLUSIONS: Venlafaxine showed some evidence of superiority to paroxetine in this difficult-to-treat patient population.