A comprehensive quantitative evaluation of the relationship between peripheral blood visfatin concentrations and coronary artery disease (CAD) is lacking. This study is the first attempt to quantify this relationship via a meta-analysis of published observational studies in terms of weighted mean difference (WMD). Literature retrieval, article selection, and data extraction were conducted. Heterogeneity was inspected using both subgroup and meta-regression analyses. In total, 15 articles involving 1053 CAD cases and 714 controls were included. Overall, peripheral blood visfatin concentrations were significantly higher in CAD cases than in controls (WMD: 4.72 ng/mL; 95% confidence interval [CI]: 2.97-6.47; P < .001), with significant heterogeneity and publication bias. Six studies were theoretically missing based on filled funnel plot, and considering the impact of these missing studies still detected a significant overall mean difference in visfatin (WMD: 2.82 ng/mL; 95% CI: 2.22-3.58; P < .001; number of studies: 21). Subgroup and meta-regression analyses indicated age, body mass index, race, diabetes, systolic blood pressure, triglycerides, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol were identified as possible causes of heterogeneity. In conclusion, our findings suggest that increased peripheral blood visfatin concentrations may be a risk marker of CAD.