Purpose: To specify the indications of surgical treatment of hypertensive intracerebral hemorrhage (ICH). Methods: During the period from Jan. 1998 to Sep. 2000, there was a prospective randomized controlled clinical trial undertaken in 17 hospitals in Shanghai. The independent correlation factors were specified according to the multivariate statistical analysis, and the outcomes of non-surgical and surgical groups were compared with each other in different strata. Results: The prognosis of ICH was based on the patient's clinical grading, volume of hematoma, consciousness, and the Glascock Coma Scale (GCS) scale. For the ICH patients of severe and moderate clinical grade, or with hematomas larger than 30 mL, the outcome of surgical group was definitely better than non-surgical group. However, for the ICH patients of mild clinical grading, or with hematomas smaller than 30 mL, there was no significantly difference in outcome between non-surgical and surgical groups. There was also no significant correlation between the outcome and the location of supratentorial hematoma. Conclusions: The study suggested the indications of surgical treatment of hypertensive intracerebral hemorrhage.