The level of serum calcium appears to be associated with blood pressure and metabolic risk factors for cardiovascular disease. Determinants of serum calcium may therefore be of interest. In a health survey in Tromsø in 1994-1995, 27,159 subjects were examined. The survey included measurements of serum calcium and questionnaires on diet and lifestyle factors. In males mean serum calcium declined from 2.41 mmol/l for those in their 20s to 2.34 mmol/l for those in their 80s. In females mean serum calcium was stable at a level of 2.35 mmol/l before the menopause, and thereafter reached a plateau of 2.39 mmol/l. In both sexes serum calcium showed a positive association with body mass index (BMI) and coffee consumption that persisted after correcting for other variables in a multiple regression model (p < 0.05). Physical activity had no significant association with serum calcium. In females alcohol consumption was negatively, and cigarette smoking positively associated with serum calcium (p < 0.01). No significant effect on the serum calcium levels was found for the intake of calcium or vitamin D, except for males with a calcium intake below 200 mg/day. Some of the observed effects, like the variation with age, may partly be explained by alterations in levels of serum albumin to which approximately 40% of circulating calcium is bound and which was not adjusted for in this study, whereas that is hardly the case for the association with BMI and coffee consumption. However, none of these factors could affect the serum calcium level more than 0.02 mmol/l, and the biological significance of the observed associations questionable.