In this article we discuss five requirements that theoretically must be fulfilled for transnatal chemosensory learning to occur in three placental species, the rat, the sheep, and the human, viz. (1) minimum or partial maturity of nasal chemoreceptor systems, (2) efficient odorivector compounds in the fetal environment, (3) the ability to memorize chemosensory information across birth, (4) perinatal continuity in chemical signals, (5) neonatal ability to detect air-borne odorants previously experienced in the aquatic environment. A substantial body of data is reviewed for the rat, in which fetal chemosensation is now firmly established. The less studied ovine perinate also shows preliminary evidence of nasal chemoreception and of postnatal retention of prenatally experienced odorants. Concerning the human species, we discuss extensive anatomical data supporting nasochemoreception in utero, but as yet no direct or indirect functional demonstration is provided. Furthermore, the strongest evidence of odorivector compounds in amniotic fluid is from human data. The results presented allow generalization of chemosensory functioning in utero in the species considered.