Objective To evaluate the efficacy and safety of percutaneous ethanol injection for hepatocellular carcinomas of 3 cm or less. Methods Randomized controlled trials (RCTs) from the Cochrane Controlled Trials Register (Cochrane Library issue 2, 2008), PubMed (1966 to 2008), EMbase (1966 to 2008), CBMdisc (1978 to 2008), and CNKI (1979 to 2008) were electronically searched. We hand searched related published and unpublished data and their references. Randomized controlled trials of percutaneous ethanol injection to treat hepatocellular carcinomas of 3 cm or less were included. Data were extracted and evaluated by two reviewers independently using a designed extraction form. RevMan 4.2.10 software was used for data analysis. Results Seven RCTs involving 891 patients were included. We conducted subgroup analyses based on outcome measures and interventions. Compared with RFA, for treatment of hepatocellular carcinomas of 3 cm or less, PEI showed statistical differences in one and three-year survival rates and one and three-year local recurrence rates after treatment(RR=0.95, 95%CI 0.91 to 1.00; RR=0.80, 95%CI 0.71 to 0.91; RR=2.18, 95%CI 1.11 to 4.30; RR=2.59, 95%CI 1.55 to 4.32). As for hepatocellular carcinomas of 2-3 cm, PEI showed statistical difference in three year cancer free survival rates after treatment (RR=0.47, 95%CI 0.24 to 0.93) .Conclusion Considering the relatively poor quality of most included trials and small sample size, insufficient evidence was obtained in this systematic review. Therefore, more randomized controlled trials with high quality are still needed to assess and verify the efficacy and safety of this treatment.