Studies of possible association between asymptomatic hyperglycemia and coronary disease were conducted among middle-aged employed men in Chicago. In one cohort, frequency of ECG abnormalities at baseline was examined among men at different glucose levels. In univariate analysis, those at the highest glucose levels had higher rates of ECG evidence of CHD, but this association did not remain when such factors as age, blood pressure, weight, cholesterol and cigarettes were taken into account in multivariate analysis. Among the studies, long-term follow-up to mortality (10, 13 and 15 yr) showed inconsistent findings on association of hyperglycemia and subsequent death. In one cohort (Gas Company men followed for 10 yr after a 1-hr post-load glucose determination) a positive glucose-mortality association was found in both univariate and multivariate analysis. But in the two remaining cohorts (Gas Company 13 yr after casual glucose determination and Western Electric 15 yr after 2-hr post-load determination), no such association was found. While differences in method and sample size as well as other limitations may have contributed to the differences observed, the inconsistency of the findings do not permit a conclusion that asymptomatic hyperglycemia is a positive coronary risk factor.