To explore the effect and safety of topical propranolol hydrochloride gel for treatment of infantile hemangioma. METHODS Thirty nude mice (BALA/c, nu/nu) were divided into three groups, experimental group, control group and normal group. Human hemangioma endothelial cells cultured in vitro were injected subcutaneously in experimental group and control group to establish infantile hemangioma model. Topical propranolol hydrochloride gel was applied on the surface of the hemangioman in experimental group and normal group. Tumor volumn change and the skin situations (edema, erythema, ulceration) were observed at different periods. 45 days after cell injection, the mice were killed and plasma concentration was detected in the experimental group and the control group by high performance liquid chromatography with evaporative light scattering detector, and tumors were subjected to histopathologic examination and immunohistochemistry for CD31 and CD34. The correlation between volumes and plasma concentration was statistically analyzed with SPSS 13.0 paired samples t test with α = 0.05 as statistical standard. At 45 days, the volume of the tumor in control group was (366.57 ± 17.08) mm³, which has a significant difference as compared to the experimental group (13.36 ± 2.09) mm³ (P < 0.05); and the plasma concentration was (16.83 ± 1.53) ng/ml in experimental group, and (18.42 ± 2.21) ng/ ml in normal group (P > 0.05 ). Topical propranolol hydrochloride gel (3%) has no irritation to nude mice's skin. Topical application of 3% propranolol hydrochloride gel is effective and safe for the treatment of infantile hemangioma.