ABSTRACT: Thiazolidinediones have shown antidepressant effect in animal studies, as well as in some uncontrolled studies evaluating human subjects with concurrent major depressive disorder (MDD) and metabolic syndrome. Although these drugs are insulin sensitizers, they also have important anti-inflammatory, neuroprotective, and anti-excitotoxic properties. Thus, we hypothesized that they would show antidepressant effect in patients with MDD even if it was not accompanied by metabolic disturbances. In this double-blind placebo-controlled study, 40 patients with MDD (DSM-IV-TR) and Hamilton depression rating scale-17 (Ham-D) score ? 22 were randomized to citalopram plus pioglitazone (15 mg every 12 h) (n=20) or citalopram plus placebo (n=20) for 6 weeks. Patients were evaluated using Ham-D (weeks 0, 2, 4, 6). Repeated-measure analysis of variance (ANOVA) and analysis of covariance were used for comparison of scores between the two groups. Treatment response (? 50% reduction in Ham-D score), remission (Ham-D score ? 7), and early improvement (? 20% reduction in Ham-D score within the first 2 weeks) were compared between the two groups using Fisher's exact test. Pioglitazone showed superiority over placebo during the course of the trial (F(1, 38)=9.483, p=0.004). Patients in the pioglitazone group had significantly lower scores at all time points than the placebo group (P<0.01). Frequency of early improvement, response (week 6), and remission was significantly higher in the pioglitazone group (95%, 95%, 45%, respectively) than in the placebo (30%, 40%, 15% respectively) group (P<0.001, <0.001, 0.04, respectively). Frequency of side effects was similar between the two groups. Pioglitazone is a safe and effective adjunctive short-term treatment in patients with moderate-to-severe MDD even in the absence of metabolic syndrome and diabetes.