Background: Anatomic resection (AR) is widely performed for hepatocellular carcinoma (HCC), but it is generally not considered superior to non-anatomic resection (NAR) in terms of prognosis. So we compared the prognosis of AR with that of NAR for HCC. Methods: We searched for articles about AR versus NAR for HCC published between January 1998 and December 2018 in PubMed, Cochrane Library, EMBASE and Wanfang database. Meta-analysis was performed on patient characteristics, tumor characteristics, operative characteristics, perioperative outcomes and long-term outcomes. Results: A total of 38 studies involving 9122 patients were included: 5062 were in the AR group and 4060 in the NAR group. Only one study included in our meta-analysis was randomized controlled trial, and others were comparative cohort studies. The AR group had an advantage over NAR group in the aspect of age, liver cirrhosis level and liver reserve function; but had a disadvantage in the aspect of tumor size, AFP level, operation time, blood loss, microvascular invasion, pathological differentiation and postoperative complication. The AR group gained 1-, 3-, and 5-year overall survival (OS) and disease-free survival (DFS) benefits versus NAR group, but there was significant heterogeneity between groups in terms of patient and tumor characteristics. Conclusion: AR is superior to NAR regarding the long-term outcomes considering the relatively acceptable heterogeneity. More prospective randomized controlled trials are required to further confirm the actual effect of AR or NAR on survival for HCC with less heterogeneity.