OBJECTIVE： To observe therapeutic efficacy and safety of Xingnaojing and naloxone hydrochloride in the treatment of coma patients with cerebral hemorrhage. METHODS： 126 coma patients with cerebral hemorrhage were randomly assigned into Xingnaojing group and naloxone hydrochloride group with 63 cases in each group. Both groups were given 20％ mannitol or and dexamethasone intravenously for dehydration therapy and other basic treatment. Xingnaojing group was given Xingnaojing injection 20 ml added into 5％ Glucose injection 250 ml， ivgtt， qd； naloxone hydrochloride group was given Naloxone hydrochloride injection 4 mg added into 5％ Glucose injection 250 ml， ivgtt， qd. Clinical efficacy， length of stay， NIHSS before and after discharge， GCS and cerebral hematoma volume before and after treatment were observed in 2 groups， and ADR was recorded. RESULTS： The total effective rate of observation group was 88.89％， which was significantly higher than that of control group （74.61％）， with statistical significance （P＜0.05）； average length of stay in observation group was （17.5±8.8）d， which was significantly lower than （28.4±10.3）d of control group， with statistical significance （P＜0.05）. NIHSS of patients before discharge and 3 months after discharge were significantly lower than before admission， and the naloxone hydrochloride group was lower than the Xinnaojing group， with statistical significance （P＜0.05）. After treatment， GCS were increased significantly and cerebral hematoma volume of 2 groups were significantly lower before treatment， and the naloxone hydrochloride group was lower than the Xingnaojing group， with statistical significance （P＜0.05）. No obvious ADR was found in 2 groups. CONCLUSIONS： Naloxone hydrochloride is better than Xingnaojing in improving neurological recovery and clinical efficacy， and shows good safety.