Sixteen volunteers aged 52-67 had their sleep recorded electrophysiologically after each of four conditions of bedtime food intake: (a) an inert capsule, (b) milk, (c) a proprietary malted milk drink, and (d) a flavored drink nutritionally equivalent to the proprietary malted milk drink. Analysis of variance revealed no significant differences over the four treatments, but the study had set out to test the specific predications that the malted milk drink would lead to longer and less broken sleep than the capsules, milk alone, or the flavored drink. In the first 6 hr of accumulated sleep, the malted milk drink was associated with significantly less broken sleep when compared with the milk or the flavored drink but not when compared with the inert capsule. The last finding differed from previous reports. To examine this difference the usual daily pattern of food intake of the subjects was assessed. The subjects were divided into two groups of 8 according to whether, at home, they usually ate a lot (Eaters) or a little food (Non-eaters) near bedtime in relation to their total intake after 1700 hr. These two groups differed in how they slept in the sleep laboratory after the four treatments. The Eaters slept significantly worse after capsules compared with milk or the malted milk drink, whereas the Non-eaters tended to sleep best after capsules, i.e., no nourishment. The subjects were ranked (1-16) according to the relative amount of food they ate in the latter part of the evening, and this rank order was found to be significantly correlated with the sleep pattern on the various sleep treatments. The more a person usually ate near bedtime, the greater the impairment of sleep after the capsules, which had no nutritive value, compared with sleep after milk or the malted drink. The smaller the amount subjects usually ate near bedtime, the better they slept after capsules compared with milk or the malted milk drink. The general conclusion from this study is that a departure from a person's usual pattern of food intake in the evening impairs subsequent sleep.