© 2016 Elsevier Ltd. All rights reserved. Background Tumor necrosis factor-α (TNF-α) is a key inflammatory mediator and its reduction is a therapeutic target in several inflammatory diseases. Curcumin, a bioactive polyphenol from turmeric, has been shown in several preclinical studies to block TNF-α effectively. However, clinical evidence has not been fully conclusive. Objective The aim of the present meta-analysis was to evaluate the efficacy of curcumin supplementation on circulating levels of TNF-α in randomized controlled trials (RCTs). Methods The search included PubMed-Medline, Scopus, Web of Science and Google Scholar databases by up to September 21, 2015, to identify RCTs investigating the impact of curcumin on circulating TNF-α concentration. Quantitative data synthesis was performed using a random-effects model, with weighed mean difference (WMD) and 95% confidence interval (CI) as summary statistics. Meta-regression and leave-one-out sensitivity analyses were performed to assess the modifiers of treatment response. Results Eight RCTs comprising nine treatment arms were finally selected for the meta-analysis. There was a significant reduction of circulating TNF-α concentrations following curcumin supplementation (WMD: -4.69 pg/mL, 95% CI: -7.10, -2.28, p <0.001). This effect size was robust in sensitivity analysis. Meta-regression did not suggest any significant association between the circulating TNF-α-lowering effects of curcumin with either dose or duration (slope: 0.197; 95% CI: -1.73, 2.12; p = 0.841) of treatment. Conclusion This meta-analysis of RCTs suggested a significant effect of curcumin in lowering circulating TNF-α concentration.