There is no consensus on the optimal operative technique for humeral shaft fractures. This meta-analysis aims to compare minimal-invasive plate osteosynthesis (MIPO) with open reduction internal fixation (ORIF) for humeral shaft fractures regarding non-union, re-intervention, radial nerve palsy, time to union, operation duration and functional outcomes. PubMed/Medline/Embase/CENTRAL/CINAHL were searched for both randomized clinical trials (RCT) and observational studies comparing MIPO with ORIF for humeral shaft fractures. Effect estimates were pooled across studies using random effects models and presented as weighted odds ratio (OR), risk difference (RD), mean difference (MD) and standardized mean difference (SMD) with corresponding 95% confidence interval (95%CI). Subgroup analysis was performed stratified by study design (RCTs and observational studies). A total of two RCT's (98 patients) and seven observational studies (263 patients) were included. The effect estimates obtained from observational studies and RCT's were similar in direction and magnitude. MIPO carries a lower risk for non-union (RD: 5%; OR 0.3, 95% CI 0.1-0.9) and secondary radial nerve palsy (RD 5%; OR 0.3, 95%CI 0.1- 0.9). Nerve function eventually restored spontaneously in all patients in both groups. Results were inconclusive regarding re-intervention (RD 7%; OR: 0.7, 95%CI 0.2-1.9), infection (RD 4%; OR 0.4, 95%CI 0.1-1.5), time to union (MD -1 week, 95%CI -3 - 1) and operation duration (MD -13 minutes, 95%CI -38.9 - 11.9). Functional shoulder scores (SMD 0.01, 95%CI -0.3 - 0.3) and elbow scores (SMD 0.01, 95%CI -0.3 - 0.3) were similar for the different operative techniques. MIPO has a lower risk for non-union than ORIF for the treatment of humeral shaft fractures. Radial nerve palsy secondary to operation is a temporary issue resolving in all patients in both treatment groups. Although both treatment options are viable, the general balance leans towards MIPO having more favorable outcomes. Copyright © 2020 Elsevier Ltd. All rights reserved.