The effects of sleeping position upon body temperature were assessed by continuous monitoring of rectal temperature in 137 babies sleeping at home under conditions chosen by their parents. There were three groups of subjects: (1) normal babies aged 12-22 weeks whose temperature rhythms were developed, (2) normal babies aged 6-12 weeks who were developing their night time temperature rhythms, and (3) babies the night after diphtheria, pertussis, and tetanus immunisation, whose temperature rhythms were disturbed. Sleeping in the prone position was not associated with higher rectal temperatures at any time of night in young babies, nor did it exaggerate the disturbance of rectal temperature rhythm after immunisation. In older normal babies the prone position did not disturb rectal temperature in the first part of the night, though prone sleepers warmed a little faster prior to walking, especially in warm conditions. Prone sleepers were, however, born earlier in gestation and tended to be of lower birth weight. Normal babies can therefore thermoregulate effectively whatever their sleeping posture, even in warm conditions, though the prone position may make it slightly more difficult to lose heat. It is difficult to see how the prone position, even interacting with warm conditions, could induce lethal hyperthermia in otherwise normal babies. Perhaps the prone position is associated with other risk factors for sudden infant death syndrome.