Since the early 1970s it has been common practice for 'rooming-in' to take place during the day in Swedish maternity wards. At night newborn babies are usually looked after by nursing staff in special nurseries. One reason for this is to avoid disturbing the mothers' sleep at night. To promote mother-infant adaptation, we undertook a programme intended to encourage night-time rooming-in. Evaluation of the programme was carried out as a quasi-experiment, divided into a pretest period (I), a 6-month implementation period, and a post-test period (II). Breast feeding and maternal sleep were studied during the first 3 postpartum days by means of self-report by 104 mothers in Period I and 111 mothers in Period II. The number of hours that the babies spent in the nursery decreased from Period I to Period II, a difference that was most obvious during the second and third postpartum nights. No difference was found in the number of breast feeds in Periods I and II, except during the third night, when Period II mothers breast fed more often. In spite of increased rooming-in in Period II, these mothers slept the same number of hours and felt equally alert as Period I mothers.