Nearly all laboratory research on human infant sleep assumes that solitary sleeping is the normal and desirable environment. Yet solitary sleeping in infancy is a very recent custom limited to Western industrialized societies, and most of the world's people still practice parent-infant cosleeping. A hypothesis is presented that cosleeping provides a sensory-rich environment which is the more appropriate environment in which to study normal infant sleep. In addition, two preliminary, in-laboratory, polygraphic investigations of mother-infant cosleeping are reported in normal infants within the peak age range for sudden infant death syndrome (SIDS). Five mother-infant pairs coslept for 1 night in the first study; in the second, three additional pairs slept separately for 2 nights and coslept the third consecutive night. The results suggest that cosleeping is associated with enhanced infant arousals and striking temporal overlap in infant and maternal arousals. Infant sleep also showed subtle alterations with cosleeping, as manifested in increased overlap with corresponding maternal sleep stages and decreased amount of Stage 3-4. These are the first in-laboratory investigations of parent-infant cosleeping. The implications of the hypothesis and preliminary results for research on the normal development of infant sleep and on SIDS are discussed.