The Hertfordshire Cohort Study based in the U.K. was the first to report associations between fetal or infant growth and the prevalence of adult glucose intolerance and diabetes. Many studies have replicated the findings with respect to birth weight, but there have been fewer observations in relationship to infant growth, because this is infrequently recorded in routine datasets. Recently, we carried out glucose tolerance tests in a more recently born group of men and women from the Hertfordshire Cohort Study. The objective was to determine whether the associations with weights at birth and 1 year of age reported in the original study of people born between 1920 and 1930 were observed in people born between 1931 and 1939. Birth weight was inversely related to the overall prevalence of diabetes (comprising newly diagnosed as well as existing cases) in men and women. However, weight at 1 year of age was not associated with diabetes in either sex. Analysis of data from the glucose tolerance tests showed that both sexes had evidence of higher insulin and glucose concentrations in people who were small at birth or during infancy. Finally, direct comparison of 2-h plasma glucose concentrations in the previous and current Hertfordshire study suggested that both surveys showed broad similarity of the trends in glucose tolerance with birth or infant weights; most differences arose at the extremes of the birth weight, possibly because of the small numbers of subjects studied in these groups.