Beta-2-microglobulin (B2M) has been suggested as an emerging biomarker for cardiovascular diseases (CVD), including coronary heart disease (CHD) and stroke, and mortality. Three databases were searched from inception to January 2, 2020, supplemented by scanning reference lists of identified studies. We identified studies that reported associations of baseline serum or plasma B2M and CVD incidence, CVD mortality, or CHD and stroke separately, in either general populations or patients with renal disease. Relative risks (RR) were extracted and harmonized to a comparison of the highest versus lowest third of the distribution of B2M, and the results were aggregated. Sixteen studies (5 in general populations, and 11 in renal disease populations) were included, involving 30,988 participants and 5391 CVD events. Based on random-effects meta-analysis, the pooled adjusted RRs comparing the highest versus lowest third of the distribution of B2M were 1.71 (95%CI: 1.37–2.13) for CVD, 2.29 (1.51–3.49) for CVD mortality, 1.64 (1.14–2.34) for CHD, and 1.51 (1.28–1.78) for stroke, with little to high heterogeneity between studies (0.0% ≤ I2 ≤ 80.0%). The positive associations between B2M and risks of CVD outcomes remained broadly significant across subgroup analyses. Moreover, the pooled adjusted RRs were 2.51 (1.94–3.26; I2 = 83.7%) for all-cause mortality and 2.64 (1.34–5.23; I2 = 83.1%) for infectious mortality. Available observational data show that there are moderate positive associations between B2M levels and CVD events and mortality, although few studies have been conducted in general populations. Display omitted •Available epidemiological evidence shows that circulating B2M levels are moderately associated with CVD events and mortality.•Future large-scale general population-based prospective studies and genetic studies are needed to assess the causality.