Abstract: Objective To systcmatically evaluate the effectiveness and safety of recombinant human growth hormone (rhGH) in treating adults with severe burn. Methods Databases including PubMed, Cochrane Library, and Embase were searched using key words " burns, thermal, human growth hormone, growth hormone, hGH, and somatropin (human)" , and China Biology Medicine disc, Chinese Journals Full-text Database, VIP Database, and Wanfang Database were searched using key words in Chinese version "烧伤,重组人生长激素" to obtain the randomized controlled trials about rhGH in the treatment of adults with severe burn from the establishment of each database to December 2016. The measurement indexes included hemoglobin (Hb) and plasma total protein, inflammatory factors [including interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α)], incidence rate of sepsis, incidence rate of hyperglycemia, wound healing time, length of stay, and mortality rate. Meta-analysis was conducted by RevMan 5.3 statistical software. Results A total of 8 trials involving 534 patients were included; 276 patients in rhGH group were treated with rhGH and 258 patients in placebo control group were treated with placebo. One trial had low risk of bias, while the other 7 trials had unclear risk of bias. The levels of Hb and plasma total protein of patients in rhGH group were higher than those in placebo control group, with standardized mean differences (SMDs) respectively 2.00 and 2.23 [with 95% confidence intervals (CIs) respectively 0.19-3.82 and 1.21-3.26, P<0.05 or P<0.01]. The levels of IL-6 and TNF-α of patients in rhGH group were lower than those in placebo control group, with SMDs respectively -1.46 and -1.13 (with 95% CIs respectively -2.40--0.53 and -1.75--0.51, P values below 0.05). Incidence rate of sepsis and mortality rate of patients in rhGH group were lower than those in placebo control group, with relative risks (RRs) respectively 0.60 and 0.35 (with 95% CIs respectively 0.42-0.85 and 0.15-0.83, P values below 0.05). Incidence rate of hyperglycemia of patients in rhGH group was higher than that in placebo control group, with RR of 2.39 (with 95% CI 1.79-3.18, P<0.001). The wound healing time and length of stay of patients in rhGH group were lower than those in control group, with SMDs respectively -1.54 and -2.00 (with 95% CIs respectively -2.22--0.86 and -3.51--0.49, P<0.05 or P<0.01). Hb, plasma total protein, inflammatory factors, incidence rate of sepsis, wound healing time, length of stay, and mortality rate showed no significant publication bias (P values above 0.05), while there may be publication bias in incidence rate of hyperglycemia (P=0.026). Conclusions rhGH can inhibit the breakdown of Hb and plasma total protein, reduce the level of inflammatory factors and incidence rate of sepsis, thus shorten the wound healing time and length of stay, thereby reduce mortality rate of adult patients with severe burn. However rhGH may cause hyperglycemia.