A Southern California community study of 4,541 men and women (aged 20-79 years) showed significant seasonal variation in fasting plasma glucose. There was a mean 0.6 mmol/l difference between highest levels in winter and lowest levels in spring. This difference was consistent over a two-year period, similar at all ages and seen in both men and women. Fasting plasma glucose levels correlated directly with percentage possible sunshine (p = 0.03) and inversely with temperature (p = 0.04). Adjustment for the level of measured obesity did not alter the observed association, but a history of recent weight gain corresponded to the season of maximum fasting plasma glucose level.