OBJECTIVE--To examine the association between non-fasting serum glucose concentrations and major ischaemic heart disease (IHD) events (fatal and non-fatal myocardial infarction). DESIGN--A prospective study. SUBJECTS--A population based sample of 7735 middle aged British men. Known diabetics, men with a glucose concentration > or = 11.1 mmol/l at screening, and hypertensive patients taking regular medication were excluded from the analysis. With exclusions (n = 509) and missing glucose values (n = 49), there were 7177 men available for analysis. MAIN OUTCOME MEASURES--Major IHD events (fatal and non-fatal myocardial infarction) during 9.5 years follow up on all men. RESULTS--There were 505 major IHD events, 222 fatal and 283 non-fatal, in the 7177 men studied. There was a non-linear relation between the glucose concentration and the risk (per 1000 men per year) of all major IHD events and fatal IHD events, with the excess risk in the upper quintile of the glucose distribution (> or = 6.1 mmol/l). The unadjusted relative risks (RR) in the upper glucose concentration quintile compared with the first to the fourth quintiles combined were 1.4 (95% CI 1.1, 1.7) for all events and 1.3 (95% CI 1.0, 1.7) for fatal events. Adjustment for age, smoking, occupational status, body mass index, physical activity, systolic blood pressure, total and high density lipoprotein cholesterol, and triglyceride concentrations had a minimal effect on these relative risk estimates. This non-linear relationship between the serum glucose concentration and the risk of a major IHD event was observed in men with no evidence of IHD at screening (n = 5518) but not in men with IHD (n = 1659). In the former group, the RR (adjusted for major coronary risk factors) for all major IHD events in the upper quintile relative to the lower quintiles combined was 1.5 (95% CI 1.2, 2.0) and for fatal IHD events was 1.8 (95% CI 1.1, 2.6). CONCLUSION--These data suggest that asymptomatic hyperglycaemia is an independent risk factor for major IHD events.