Sudden infant death syndrome (SIDS) is the leading cause of human infant mortality after the neonatal period in Western countries. Recently, child care practices have been shown to be important in determining infant vulnerability to SIDS. However, very little is known about the impact of parent-infant cosleeping on infant sleep physiology and behavior and SIDS risk. This reflects the failure of Western societal research paradigms to appreciate the human infant's evolutionary history of cosleeping, the recency of the emergence of solitary infant sleeping as a practice and the fact that parent-infant cosleeping is still the preferred sleeping arrangement for the majority of contemporary societies. Incorporating current hypotheses on the mechanisms of SIDS, we have hypothesized that the comparatively sensory-rich cosleeping environment might be protective against SIDS in some contexts. As a first step to characterize cosleeping environments, this investigation is aimed at assessing, in routinely bedsharing mothers and infants, their relative sleeping positions and the potential for sleeping in close face-to face proximity and for infant exposure to increased environmental CO2 produced by maternal respiration. The latter is important in that breathing elevated levels of CO2 can have diverse effects, ranging from respiratory stimulation at low levels to suffocation at very high levels. Two related laboratory studies were performed. In the first, all-night videotapes of 12 healthy, routinely bedsharing mother-infant pairs were analyzed for sleeping positions and time spent in face-to-face orientation and distances separating their faces. Infants were 11-15 wk old. Mothers predominantly positioned themselves on their sides facing their infants, with the infants placed either supine or on their sides. Mothers and infants slept oriented face-to-face for 64 +/- 27% (S.D.) of non-movement time, with distance less than 20 cm commonly separating their faces. In the second study, concentrations of CO2 in air were measured in six young women at distances of up to 21 cm from their nares. Peak expiratory CO2 concentrations remained above 1.0% at distances up to 9 cm and above 0.5% at 18 cm. Both baseline and peak CO2 levels were further increased at all distances when measured within a partial air pocket created to simulate a bedding environment sometimes seen during bedsharing. We conclude that during bedsharing there is potential for 1) a high degree of face-to-face orientation and close proximity and consequently 2) increased environmental CO2, as a result of maternal respiration, to non-lethal levels that might stimulate infant respiration. The close proximity would also maximize the sensory impact of the mother on the infant through other modalities. We also suggest that bedsharing may minimize prone infant positioning, a known risk factor for SIDS.