Medically unexplained symptoms (MUS) commonly present across the board in medical specialties and are often challenging to treat. Our objective was to assess the efficacy for cognitive-behavior therapy (CBT) in MUS. Electronic search of databases was carried out for published controlled trials in English language peer-reviewed journals from inception till August 2016. Effect sizes for the trials were computed using standardized mean difference, and I(2) test was used to assess sample heterogeneity. Pooled mean effect sizes were derived using a random-effects model. Critical appraisal of studies was done using the Cochrane risk of bias assessment tool. A total of 11 trials involving 1235 subjects were included in the study. Ten trials used standard CBT techniques while one studied the efficacy of mindfulness-based CBT technique. The control arms were treatment as usual in five trials, augmented care in four and waitlisted controls in two trials. The pooled mean effect size for CBT was 0.388 (range 0.055-0.806, 95% confidence intervals 0.316-0.461). The I(2) value was 0 using a random effects model indicating low heterogeneity among studies. Risk of bias was noted in many included studies. Egger plot intercept indicated potential publication bias. CBT was superior to the waiting list, treatment as usual or enhanced usual care with moderate effect sizes in the treatment of MUS. These findings are impacted by the limited number of studies in this area and questionable methodological rigor of included studies.