OBJECTIVE To systematically review the clinical efficacy and safety on TCHMs that are used for vascular dementia(VaD).METHODS To identify studies for systematical review,electronic searches were performed through several databases-ALOIS,CNKI,CBM,Weipu,Wanfang,etc.Only randomized control trials(RCTs)or controlled clinical trials(CCTs)were included.Patients were diagnosed with VaD by diagnostic criteria(DSM,NINDS-AIREN,ICD or HIS)as well as imaging technique(CT,MRI or functional imaging,etc).Eligible TCHMs must be recognized in the Chinese Pharmacopeia or the National Essential Drug List of the People′s Republic of China.Included studies were appraised using the Cochrane Collaboration risk-of-bias criteria.Efficacy and safety outcomes were evaluated by meta-analysis.Efficacy outcomes include cognition,daily function,global performance and behaviour;safety was assessed by the number of adverse events and number of subjects experiencing adverse events.Assessment of heterogeneity,subgroup analysis and sensitivity analysis were also performed.RESULTS A total of 46 trials on 29TCHMs(3522patients)were included.45 studies were RCTs and 1was CCT.In these 45 RCTs,only 2were appraised as adequately randomised.5of 46 trials were appraised as having low risk of bias in blinding.Sample sizes were generally small ranging from 26 to 216with a median of 68.All trials were conducted in China from 1997 to 2013.All 46 studies assessed cognition using one or a combination of the following scales:MMSE(n=40 studies;3096 patients),HDS(n=22;1664 patients),ADAS-Cog(n=4;241 patients),CDT(n=1;60patients)and CCSE(n=1;26patients).Half of the studies assessed daily function using either the ADL(n=22;1743 patients)or IADL(n=2;203patients).Only 6studies measured behaviour using the FAQ(n=3;226 subjects),BBS(n=1;48patients),NPI(n=1;100subjects)or Neurological Deficits Function Scale(n=1;91patients).30 studies measured global performance.31 of 46 studies made conclusions regarding the safety of the TCHMs.Despite the problems of methodology and reporting,we can identify three TCHMs-NaoXinTong,Shenfu and Tongxinluo as having relatively stronger evidence of efficacy.There is weak evidence for the safety of TCHMs for VaD.CONCLUSION There is weak evidence for the efficacy and safety of TCHMs for VaD because of the poor methodology,short duration of follow-up and inadequate reporting.However the agents appear to be relatively free of severe short-term AEs,hence we encourage further better designed and reported trials. OBJECTIVE To systematically review the clinical efficacy and safety on TCHMs that are used for vascular dementia (VaD). METHODS To identify studies for systematical review, electronic searches were performed through several databases - ALOIS, CNKI, CBM, Weipu, Wanfang, etc. Only randomized control trials (RCTs) or controlled clinical trials (CCTs) were included. Patients were diagnosed with VaD by diagnostic criteria (DSM, NINDS-AIREN, ICD or HIS) as well as imaging technique (CT, MRI or functional imaging, etc) .Eligible TCHMs must be recognized in the Chinese Pharmacopeia or the National Essential Drug List of the People’s Republic of China.Included studies were appraised using the Cochrane Collaboration risk-of-bias criteria. Efficacy and safety outcomes were evaluated by meta-analysis .Efficacy statements include cognition, daily function, global performance and behavior; safety was assessed by the number of adverse events and number of subjects experiencing adverse events. Assessment of heterogeneity, subgroup analysis and sensitivity analysis were also performed .RESULTS A total of 46 trials on 29TCHMs (3522patients) were included.45 studies were RCTs and 1was CCT. These 45 RCTs, only 2were appraised as adequately randomized.5of 46 trials were appraised as having low risk of bias in blinding. Sample sizes were generally small ranging from 26 to 216 with a median of 68. All trials were conducted in China from 1997 to 2013.All 46 studies assessed of using one or a combination of the following scales: MMSE (n = 40 studies; 3096 patients), HDS (n = 22; 1664 patients), ADAS-Cog (n = 4; 241 patients), CDT (n = 1; 60patients) and CCSE the studies assess daily using either the ADL (n = 22; 1743 patients) or IADL (n = 2; 203 patients) .Only 6studies measured behavior using the FAQ (n = 3; 226 subjects) ), NPI (n = 1; 100subjects) or Neurological Deficits Function Scale (n = 1; 91patients) .30 studies measured global performance.31 of 46 studies made conclusions regarding the safety of the TCHMs.Despite the problems of methodology and reporting, we can identify three TCHMs-NaoXinTong, Shenfu and Tongxinluo as having weak stronger of of efficacy. There is weak evidence for the safety of TCHMs for VaD .CONCLUSION There is weak evidence for the efficacy and safety of TCHMs for VaD because of the poor methodology, short duration of follow-up and inadequate reporting. Despite the agents appear to be relatively free of severe short-term AEs, therefore we even better better designed and reported trials.