Aims: To investigate the associations between metabolic score for visceral fat (METS-VF) and clinical outcomes among populations with different glucose tolerance statuses. Methods: We analysed 6827 participants aged ≥ 40 years with different glucose tolerance statuses from a cohort study. The associations between METS-VF and cardiovascular disease (CVD) events and all-cause mortality were assessed using Cox regression, restricted cubic spline and receiver operating characteristic curves. Results: During a follow-up of 5.00 years, there were 338 CVD events and 307 subjects experienced all-cause death. The METS-VF quartile (Quartile 4 versus 1) was significantly related to CVD events [adjusted HRs and 95% CIs: 5.75 (2.67–12.42), 2.80 (1.76–4.48), and 3.31 (1.28–8.54) for subjects with normal glucose tolerance, prediabetes and diabetes, respectively] and all-cause mortality [adjusted HRs and 95% CIs: 2.80 (1.43–5.49), 4.15 (2.45–7.01), and 4.03 (1.72–9.42), respectively]. Restricted cubic spline suggested a dose–response association of METS-VF with the risk of CVD events and all-cause mortality. The area under curve for CVD events and all-cause mortality was higher for METS-VF than for the other obesity and IR indexes in subjects with different glucose tolerance statuses. Conclusions: The METS-VF was associated with an increased risk of CVD events and all-cause mortality and could be used as a predictive index of the risk of CVD events and all-cause mortality among populations with different glucose tolerance statuses.