To determine the effect of parity on the later development of non-insulin-dependent diabetes mellitus or impaired glucose tolerance, we studied a population-based sample of 1186 women at least 40 years of age; those who had been given a diagnosis of diabetes mellitus before the age of 40 or who had insulin-dependent diabetes mellitus were excluded from the study. On the basis of the World Health Organization's criteria, 714 had normal glucose tolerance, 326 had impaired glucose tolerance, and 146 had non-insulin-dependent diabetes mellitus (NIDDM). After adjustment for age, obesity, and family history of diabetes, increased parity was associated with a significantly increased risk of both NIDDM (odds ratio, 1.16 [95 percent confidence interval, 1.04 to 1.29] per pregnancy) and impaired glucose tolerance (odds ratio 1.10 [95 percent confidence interval, 1.01 to 1.19] per pregnancy). Obesity, whether estimated by means of the body-mass index or the waist-hip ratio, was significantly associated with an increased risk of both NIDDM and impaired glucose tolerance, but this factor did not explain the association between parity and diabetes or impaired glucose tolerance; neither the maximal lifetime body-mass index nor the waist-hip ratio was significantly associated with parity in this cohort. We conclude that there is a slight increase in the risk of NIDDM or impaired glucose tolerance with increasing parity many years after childbearing and that this association is not explained by obesity.