BACKGROUND : Herbal products have been widely used as a means of ethnomedicine worldwide. Recently, the potential hepatotoxicity of herbs has become a medical issue but comprehensive studies are limited. OBJECTIVE : This study aims to determine the clinical features of herb induced liver injury (HILI) including its constituent ratio among liver injury case cohorts that included both HILI and drug induced liver injury (DILI). METHODS : A systematic review was conducted using a literature search for DILI/HILI in seven electric databases including PubMed, Cochrane and Embase. We analyzed the DILI/HILI cases and clinical characteristics in terms of herbs, conventional drugs, concomitant, or others. RESULTS : Thirty-one studies met the necessary criteria and included 9 prospective and 22 retrospective studies. Among total number of overall DILI/HILI cases (7511, male 2819, female 3669 and unknown 1023), 25.0% (1874 cases) were implicated in herbs. HILI was relatively higher in females (69.8% vs. 30.2% male), compared to conventional drugs (57.3% female vs. 42.7% male, p < 0.01), while it was prone to induce hepatocellular injury (hepatocellular 78.8%, cholestatic 8.9%, mixed type 12.3%), contrary to conventional drugs (hepatocellular 56.7% vs. cholestatic 22.1% vs. mixed 21.2%), respectively (p < 0.01). The main herbs causing HILI included Polygonum multiflorum, Psoralea corylifolia, Corydalis yanhusuo, and Rheum officinale. CONCLUSIONS : This review created the comparative and comprehensive feature of hepatotoxicity by herbal products, which provides reference data for the clinical applications and establishing pharmacovigilance system of herbs.