Introduction: The efficacy of resistance training to treat non-alcoholic fatty-liver disease (NAFLD) remains controversial. In this study, a systematic review and meta-analysis was performed to explore the influence of resistance training on NAFLD. Methods: PubMed, EMbase, Web of science, EBSCO, and Cochrane library databases were search for dates through March 2018 for randomized controlled trials (RCTs) assessing the effect of resistance training on liver function of NAFLD. Meta-analysis was performed using the random-effect model. Results: Four RCTs involving 133 patients are included in the meta-analysis. Overall, compared with control group (i.e. usual activities or aerobic activities) for NAFLD, resistance training can significantly reduce alanine transferase (ALT) (mean difference (MD) =-6.63; 95% confidence interval (CI) =-12.87 to-0.40; P = 0.04), aspartate transferase (AST) (MD =-1.17; 95% CI =-1.82 to-0.52; P = 0.0004), gamma-glutamyl transferase (GGT) (MD =-2.87; 95% CI =-5.20 to-0.54; P = 0.02), cholesterol (MD =-8.76; 95% CI =-13.08 to-4.44; P < 0.0001), low-density lipoprotein (LDL) (MD =-5.06; 95% CI =-6.25 to-3.87; P < 0.00001), but has no remarkable influence on glycosylated hemoglobin (HbA1C) (MD =-0.01; 95% CI =-0.08 to 0.06; P = 0.81). Conclusions: Resistance exercise has an important ability to reduce ALT, AST, GGT, cholesterol, and LDL in patients with NAFLD, but shows no influence on HbA1C.