Utilizing their differential susceptibilities to inhibitors and heat, we determined the amounts of the placental, liver, and fetal-intestinal isoenzyme forms of alkaline phosphatase in 143 samples of normal amniotic fluid obtained at 14 to 18 weeks' gestation (1). For reliable results, it was necessary to standardize inhibition profiles for each pure isoenzyme in amniotic fluid. Total activity and the absolute amounts of placental and fetal-intestinal activities were significantly related to gestational age (p less than 0.05). These relationships that were absent when activities were expressed as percentages of the total. The mean isoenzyme composition of the 143 samples, expressed as a percentage of total alkaline phosphatase activity, was: placental, 3.4%; liver, 9.8% (maximum, 47%); and fetal intestinal, 87% (minimum, 53%). The presence of phosphate in the assay medium (13.5 mmol/L) profoundly and differentially inhibited the isoenzymes of alkaline phosphatase and changed the inhibition profiles of the tissue-specific enzymes; thus, it would not be feasible to use inhibitors to differentiate the forms. We therefore propose a phosphate-free technique for quantifying the isoenzymes of alkaline phosphatase in amniotic fluid obtained at 14 to 18 weeks' gestation, to achieve the highest predictive values in a prenatal diagnostic test for cystic fibrosis.