Hyperimmune absorbed rabbit antisera which were reactive with epitopes specific for individual variants of human placental alkaline phosphatase were tested for their reactivity with primate placental alkaline phosphatases. Using the three epitope-specific reactivities defined previously, we found that: epitope I is present in the S-, D- and I-variants of human placental phosphatase, and in the chimpanzee and pygmy chimpanzee placentae; epitope II is present in the F- and 17-variants, and in the Nagao isoenzyme of human placental alkaline phosphatase, and in some orangutan placentae and all spider monkey placentae tested; epitope III is present in the F- and 17-variants, and the Nagao isoenzyme of human placental alkaline phosphatase, and in all the spider monkey placentae and the single squirrel monkey placenta examined. The binding assay was complemented by a competitive radioimmunoassay, which confirmed that the spider monkey placental samples were binding to the same antibody population which bound the human enzymes. The presence of epitopes characteristic of rare human placental phosphatase variants in these remote primate relatives suggests that the rare variants in the current human population have been present during the entire course of evolution. The presence of both epitopes characteristic of the Nagao isoenzyme in spider monkeys suggests that this variant isoenzyme is closely related to the enzyme present in the primate placenta at the time of species divergence (humans and New World monkeys). A hypothetical scheme for this divergence is proposed.