Background: This study was designed to evaluate the effect of melatonin on nonalcoholic fatty liver disease (NAFLD) in compared to placebo. Materials and Methods: A total of 100 patients with histopathological diagnosis NAFLD in two groups of case and control received oral melatonin or placebo thrice daily for 3 months. Collected data were weight, waist, systolic blood pressure (SBP), diastolic blood pressure (DBP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), high sensitive C-reactive protein (hsCRP), fatty liver grade, and side effects which were measured at baseline and after treatment period using standard clinical chemistry techniques. Results: Before treatment the mean of weight, waist, SBP, DBP, ALT, AST, and hsCRP between cases and controls were similar (P > 0.5). After treatment, only the differences in the mean of hsCRP in cases was significantly lower than controls (P = 0.003). In case group, all variables after treatment were significantly decreased compare to baseline (P > 0.5) and only AST after treatment was similar to before treatment (P > 0.5). The mean of a decrease in the level of weight, waist, SBP, and ALT were not statistically significant between groups (P > 0.5). In the case group in compare to control group the level of DBP, AST, and hsCRP significantly more decreased. After treatment fatty, liver grade was statistically improved in more cases than controls (P = 0.001). Side effects were similar between the two groups. Conclusion: Melatonin significantly decreases liver enzymes, so the use of melatonin in patients with NAFLD can be effective.