Objective To assess the clinical efficacy and safety of proton pump inhibitor （PPI） and H2RA for stress ulcer bleeding in stroke patients. Methods Randomized controlled trials （RCT） were identified from MEDLINE （ 1966- Oct. 2005 ） ,EMBASE （ 1984- Oct. 2005 ）, The Cochrane Library （ Issue 4,2005 ）, CBMdisc （ 1980- Oct. 2005 ） and VIP（ 1980- Oct. 2005 ）. We handsearched the related published and unpublished data and their references. The quality of included trials was evaluated. Data were extracted by two reviewers independently with a designed extraction form. RevMan 4. 2.7 software was used for data analysis. Results Twenty RCT were included with 2 624 patients. The results of meta-analysis were listed as follows： （1） stress ulcer bleeding （SUB） ： PPI （ OR 0.14,95% CI 0.08 to 0.24, NNT = 3 ） and H2RA （OR 0.24,95% CI 0.15 to 0.39, NNT =5） significantly reduced the incidence of SUB in comparison with control group. PPI significantly reduced the incidence of SUB compared with H2R.A（P 〈0. 00001 ）. （2） Mortality： PPI （OR 0.22,95% CI 0. 11 to 0.47, NNT =8） and H2RA （OR 0.53,95% CI 0. 34 to 0.81, NNT =16） significantly decreased the mortality compared with non-prophylaxis group. PPI significantly decreased the mortality compared with H2RA （OR 0.28,95% CI 0.09 to 0. 89）. （3） Adverse effect： There were not evident adverse effects in both PPI and H2RA groups. Conclusions PPI and H2RA may reduce the incidence and mortality of SUB in stroke patients, and PPls are better in reducing incidence of SUB than H2RA.