OBJECTIVES: Long-term studies of rosiglitazone (ADOPT) and pioglitazone (PROactive) have reported average weight gain of 4.8 kgs and 3.6 kgs, respectively. This study assessed weighted average gain in the body weight of type 2 diabetic patients due to the use of thiazolidinediones (TZDs) over a short duration. METHODS: Meta-analysis of published rosiglitazone and pioglitazone clinical trials with duration up to 52 weeks. Initially, trials with placebo as well as active comparator arms were identified through a systematic search of PubMed, EBSCO and Sci-lit databases. Each trial arm was treated as an independent observation. Wherever possible, placebo-subtracted weight change was computed for each arm of TZD trials. Exploratory regression analysis was conducted using backward elimination method to identify predictors of weighted average gain in the body weight. RESULTS: Overall, 43 TZD trials yielded 60 trial arms which represented 8,322 patients. Out of these, only 13 trial arms provided information sufficient to compute the placebo-subtracted weight gain. Weighted placebo-subtracted gain in the body weight due to TZDs was 3.22 ± 7.16 kg. Significant predictors (b ± SE) of placebo-subtracted weight gain (adjusted R2= 0.73) were the use of TZDs as monotherapy (1.04 ± 0.36), duration of diabetes in the treatment group in years (0.25 ± 0.06) and rosiglitazone dose of 8 mg (1.36 ± 0.26). CONCLUSION: Patients using TZDs recorded mean weight gain of 3.22 ± 7.16 kg during short-term placebo-controlled clinical trials. Duration of diabetes, use of TZD as monotherapy and rosiglitazone dose of 8 mg were the only significant predictors of placebo-subtracted gain in the body weight. This weight gain is highly undesirable in diabetics, due to high risk of experiencing a cardiovascular event and potential adherence problems.