Using the technique of centrifugal ultrafiltration isodialysis to measure the free concentration of 1,25-dihydroxyvitamin D [1,25-(OH)2D], we determined the affinity of serum proteins for 1,25-(OH)2D both by Scatchard analysis (increasing ligand concentration at fixed binding site concentrations) and by a novel analysis in which the binding site concentrations were varied (serial dilution) at fixed ligand concentrations. The high affinity binding constant in serum for 1,25-(OH)2D was 3.7 X 10(7) M-1 by Scatchard analysis and 4.2 X 10(7) M-1 by serial dilution analysis. Human serum albumin had a much lower affinity for 1,25-(OH)2D (5.4 X 10(4) M-1). When vitamin D-binding protein (DBP) was selectively removed from serum by an actin affinity column, the affinity of the remaining serum proteins for 1,25-(OH)2D was that of albumin. Postulating a two-site model (DBP and albumin) for transport of 1,25-(OH)2D in serum and incorporating the estimated affinity constants of DBP and albumin for this metabolite, we calculated that 85% of total circulating 1,25-(OH)2D is transported in blood bound to DBP in normal individuals (0.4% is free and 14.6% is bound to albumin). In patients with liver disease, 73% is bound to DBP (1.1% is free and 25.9% is bound to albumin). Using this same two site model, we found a reasonable correlation (r = 0.612; P less than 0.001) between the measured free 1,25-(OH)2D level and the calculated free 1,25-(OH)2D level in serum based on albumin and DBP concentrations in 16 normal subjects and 16 patients with liver disease. These results confirm the concept that although DBP is the principal protein carrier of 1,25-(OH)2D in serum, albumin is a major secondary carrier, especially in patients with low DBP levels.