Community pharmacy services (CPS) have been shown to be positive in many disease management and patient care programs, but clinical outcomes were followed by process indicators and methodological flaws in previous researches made it difficult to prove the effectiveness of clinical outcomes of CPS. Therefore, this study attempted to review the clinical outcomes of CPS. Interventions included are provision of medication review, patient education, adherence assessment, health/lifestyle advice, physical assessment, monitoring, prescribing, or adjusting and administering therapy from community pharmacists. By searching for key words like community pharmacists, pharmaceutical services, clinical outcomes in MEDLINE and EMBASE and manually searching (up to June 2017), 1910 studies investigating the clinical outcomes of CPS were obtained. After screening the titles, abstracts and full texts for relevancy, 52 researches with controlled groups were included and assessed for methodological quality. Finally, 25 studies were selected for the meta‐analysis based on their common endpoints: systolic blood pressure, diastolic blood pressure and glycosylated haemoglobin. The Cochrane tool was used to assess the risk of bias. Chi‐square and I‐square tests were performed to assess heterogeneity, and the weighted mean differences were estimated using random effect models. Of the 52 articles, 47 studies demonstrated that CPS had positive clinical outcomes, 3 studies showed mixed outcomes and 2 studies revealed no effects. In the meta‐analysis, intervention groups displayed greater reductions in systolic BP (95% CI: −8.198–2.356), diastolic BP (95% CI: −3.648–0.645) and HbA1c (95% CI: −0.905–0.224) than usual care groups. CPS have positive clinical outcomes, particularly significant reductions in systolic BP, diastolic BP and HbA1c. It was difficult to find out which intervention(s) of CPS directly led to certain changes and influence of CPS might be underestimated for only three common surrogate endpoints. More researches should be conducted with sufficient data.