Putative standards of skeletal alkaline phosphatase (ALP) (from bone, bone cells, osteosarcoma cells, and Pagetic serum) and hepatic ALP (from cholestatic serum and bile) were used to compare three methods for quantifying skeletal ALP activity in serum: heat inactivation, precipitation with wheat germ agglutinin (WGA), and precipitation with concanavalin A (Con A). All the skeletal ALP standards were similarly sensitive to heat inactivation, as were the hepatic ALP standards. Heat inactivation separated skeletal from hepatic ALP by a 50% difference in remaining ALP activities (e.g., 23% and 74% remaining skeletal and hepatic ALP activities after 30 min at 52 degrees C). Differential precipitations with WGA and with Con A were less efficient at separating skeletal from hepatic ALP (maximum differences of < 30% remaining ALP activity). Although both types of hepatic ALP standard (cholestatic serum and bile) were precipitated with similar efficiencies by WGA and Con A, the skeletal ALP standards were not (e.g., at 2.7 g/L, WGA precipitated 78-86% of the ALP activity in Pagetic serum, but only 49% of the ALP activity in extracts of human bone). These data suggest that heat inactivation is preferable to precipitation with WGA or Con A for quantifying skeletal ALP activity in serum: it better separates skeletal from hepatic ALP activity and is not sensitive to glycosyl heterogeneity.