Placental hormones are produced by one genetic individual (the fetus) to act on the receptors of another genetic individual (the mother). Mothers are probably able to extract some information from placental hormones, but this information may be limited to a crude measure of fetal vigor. Placental hormones are most easily interpreted as fetal attempts to manipulate maternal metabolism for fetal benefit. An evolutionary model is presented for a hypothetical hormone that increases the nutrient content of maternal blood. The model predicts that, at an evolutionary equilibrium, the hormone will be produced solely by the mother or solely by the placenta, but not by both. If the gene for the hormone is subject to genomic imprinting, the paternally-derived allele will be active and the maternally-derived allele will be silent. Hormone production benefits the members of the mother's current litter at some cost to future litters. Therefore, paternity changes between litters increase the level of hormone production. On the other hand, offspring that produce less of the hormone than litter-mates share the benefits but have lower costs. Therefore, multiple paternity within litters reduces the level of hormone production.