Indexes of retinopathy and nephropathy were studied in a large population of diabetics using standardized methods. In each of 973 subjects, more than 100 other variables were also measured: this made It possible to take into account many confounding factors when we examined the relationships of these variables to microangiopathy. The most powerful risk factor for microangiopathy was duration of diabetes, but frequency of both retinopathy and nephropathy was impressively related to the level of plasma glucose at the time of examination. This latter relationship was independent of the effects of other variables. Risk of retinopathy was also related to factors that were probably associated with higher previous plasma glucose levels—need for insulin therapy, history of ketonuria, present level of plasma triglyceride, leanness, and younger age of onset of diabetes. In subjects with a systolic blood pressure of less than 170 mm Hg and who did not have proteinuria, no significant relationship was found between blood pressure and frequency of retinopathy, but, when the systolic pressure was greater than 169 mm Hg, rates of retinopathy were excessive, even In those subjects without proteinuria. Several other factors had no significant relationship to retinopathy or nephropathy: these included age, level of education, smoking, and plasma cholesterol. Rates of severe nephropathy were somewhat greater In men than in women, but this difference was not statistically significant. Rates of retinopathy and of mild proteinuria were similar in men and women. After adjustment for duration of diabetes, the risk of retinopathy was no higher in the 102 subjects with two diabetic parents than in the 254 who had no family history of diabetes.