Calcium homeostasis was investigated in male BB rats with a diabetes duration of 3-4 weeks and compared with that in nondiabetic littermates either fed ad libitum or receiving selective semistarvation or an oral Ca supplement to obtain additional weight-matched and Ca intake-matched control groups. Diabetic rats had markedly increased food and Ca intake, so that their net Ca balance remained positive despite a 13-fold increase in urinary Ca excretion and a disappearance of active duodenal Ca absorption. Decreased duodenal Ca uptake correlated with decreased 1,25-(OH)2D3 levels (89 +/- 15 vs. 160 +/- 13 pg/ml in nondiabetic rats), decreased duodenal 9K Ca-binding protein concentrations (10 +/- 1 vs. 21 +/- 2 micrograms/mg protein), and decreased number of 1,25-dihydroxyvitamin D3 [1,25-(OH)2D3]-binding sites in duodenum, although the binding affinity was above normal. Nondiabetic Ca-supplemented rats exhibited a similar response: decreased 1,25-(OH)2D3 (95 +/- 8 pg/ml) and 9K Ca-binding protein (7 +/- 0.5 micrograms/mg protein) concentrations, decreased active duodenal Ca uptake, increased urinary Ca excretion, and a normal net Ca balance. Plasma vitamin D-binding protein levels were decreased by 62% in diabetic rats, due to a marked decrease in production rate, while the plasma half-time remained normal. The free 1,25-(OH)2D3 index was highest in diabetic rats, suggesting partial vitamin D resistance at the duodenal level. In semistarved rats, 1,25-(OH)2D3 levels and active Ca uptake remained normal, and the free 1,25-(OH)2D3 index was increased, together with suppressed vitamin D-binding protein levels. These studies indicate that nutritional abnormalities may contribute to but cannot totally explain the disturbances in vitamin D metabolism, transport, or action at the intestinal level.