Long-term exposed to ambient particulate matter (PM) has been recognized as a risk factor for cardiovascular diseases (CVDs) mortality but few studies examine the utility of WHO's interim targets (2006)concerning PM control on CVD mortality. This review aimed to synthesize the long-term exposed to ambient PM exposures on overall CVD mortality according to WHO's interim targets; meanwhile, subgroup analysis by ethnicity, smoking status and PM assessment method were also conducted. We systematically searched studies published between January 1974 and Jul 2017 in PubMed and Embase.Quality of each study was assessed using Critical Appraisal Skill Programme (CASP) checklist. Fixed effects or random-effects model of meta-analysis was determined by the test of heterogeneity. Subgroup analyses were conducted according to ethnicity, smoking status, PM assessment method and interim PM targets of WHO guidelines. Overall, 16 eligible studies were included, covering 15,511,997 participants and 542,991 CVD deaths. Five studies concerning PM2.5 were rated as good quality. The pooled hazard ratio (HR) of every 10mg/m3 increment of PM2.5 exposure for CVD mortality was 1.12 (95% CI ¼ 1.08-1.16), but it was not significant for PM10 (HR ¼ 1.02, 95%CI ¼ 0.89-1.16). Compared with the annual PM2.5 exposure level within WHO's interim targets (1.11-1.16), significantly smaller HR was observed for subjects with an exposure level below WHO's air quality guideline (HR ¼ 1.03, 95%CI ¼ 1.02-1.04). The pooled HR was relatively higher for studies in Asian and with at least 11 years' follow-up and those adopting relatively poor methods (category 1) in assessing PM2.5, whilst the risk was similar regardless of smoking status. Egger and Begg's tests showed no evidence for publication bias. Long-term ambient PM2.5 exposure level was positively associated with the overall CVD mortality. Different interim targets above the WHO's Air Quality Guideline level exerted a similar scale of CVD risk, but there is no evidence for a threshold.