A previous report suggested that 5 weeks of continued treatment with 20 mg of fluoxetine was approximately as effective as double-blind titration to a dose of 60 mg in patients who had failed to respond to 3 weeks of initial treatment at 20 mg. The current study was undertaken to evaluate whether 150 mg of sertraline was any more effective than 50 mg in treating depressed patients who were non-responders at 3 weeks. Ninety-one outpatients with DSM-IV major depressive disorder who had a 17-item Hamilton Depression Rating Scale (HAM-D) score > or = 18 were treated with open label sertraline for 3 weeks. Patients who did not achieve remission (defined as 17-item HAM-D total score < or = 8 by week 3) were then randomized to 5 more weeks of double-blind treatment with either 50 mg of sertraline or immediate titration to 150 mg of sertraline. Efficacy was assessed at each visit with the HAM-D, Clinical Global Impressions (CGI)-severity and improvement scale, and the Hopkins Symptom Checklist. There were no significant between-group differences in clinical or demographic features at baseline for the three treatment groups. After 3 weeks of open-label treatment, 16 patients were not randomized, of whom 11 (69%) met responder criteria. The remaining patients were randomized, double-blind, to 50 mg of sertraline (n = 37; HAM-D = 19.2 +/- 5.0) or 150 mg of sertraline (n = 38; HAM-D = 18.4 +/- 5.0). PROC-Mixed analyses found no significant difference in slopes for any outcome measure when comparing 50 mg and 150 mg sertraline treatment groups. At week 8 (LOCF), the overall remission rate (HAM-D < or = 8) for 3-week non-responders was 40%, with no statistically significant between-group difference for the 50 mg versus 150 mg doses of sertraline (P > 0.10). A completer analysis yielded similar results. Adverse events were mostly mild on both doses of sertraline and led to few treatment discontinuations. The results suggest that for most patients continued treatment with 50 mg dose of sertraline yields a rate of antidepressant response that is comparable to what is achieved by dose escalation from 50 mg to 150 mg of sertraline after 3 weeks of treatment. While some patients clearly benefit from higher doses, the results of the current study are consistent with the lack of any evidence for a dose-response curve with sertraline in the treatment of depression.