The purpose of this paper is to evaluate the efficacy and safety of recombinant human hepatocyte growth factor (rh-HGF) for liver failure (LF) using meta-analysis of data from the literature involving available randomized controlled trials of rh-HGF plus comprehensive therapy (CT) compared with that of CT alone (Therapy I versus II) in treating LF. We searched the Cochrane Library, MEDLINE, EMBASE, CBMdisc, and CNKI as well as employing manual searches. Based on our search criteria, we found 21 trials, involving 5902 patients. Our results showed that Therapy I, compared with therapy II, significantly reduced the overall mortality (RR=0.62; 95% CI, 0.59-0.66; p=0.0001). Compared to two clinical types of LF (acute and acute-on-chronic), therapy I perhaps had significant effect on mortality due to sub-acute LF, RR and 95% CI were 0.76 [0.65, 0.89], 0.66 [0.60, 0.74], and 0.58 [0.53, 0.64], respectively. Additionally, there was a reduction in mortality of patients that had evidence for an early stage of LF compared to the two other clinical stages of LF (Middle and Advanced); RR and 95% CI were 0.34 [0.24, 0.49], 0.49 [0.44, 0.55], and 0.87 [0.82, 0.93], respectively. No serious adverse events were reported. We conclude that Therapy I may reduce mortality in LF, especially in sub-acute LF and the early stage of LF. However, considering the strength of the evidence, additional randomized controlled trials are needed before Therapy I can be recommended routinely.