The present study aimed to evaluate evidence on metabolisable energy consumption and pattern of consumption with age in infants in the developed world who were exclusively breast-fed, at around the time of introducing complementary feeding. We carried out a systematic review aimed at answering three questions: how much milk is transferred from mother to infant?; does transfer increase with the age of the infant?; and what is the metabolisable energy content of breast milk? Thirty-three eligible studies of 1041 mother--infant pairs reported transfer at 3--4 months of age, the weighted mean transfer being 779 (SD 40) g/d. Six studies (99 pairs) measured transfer at 5 months, with a weighted mean transfer of 827 (SD 39) g/d. Five studies (72 pairs) measured milk transfer at 6 months, reporting a weighted mean transfer of 894 (SD 87) g/d. Nine longitudinal studies reported no significant increases in milk transfer after 2--4 months. Twenty-five studies on breast-milk energy content were based on 777 mother--infant pairs. The weighted mean metabolisable energy content was 2.6 (SD 0.2) kJ/g. Breast-milk metabolisable energy content is probably lower, and breast-milk transfer slightly higher, than is usually assumed. Longitudinal studies do not support the hypothesis that breast-milk transfer increases markedly with age. More research on energy intake in 5--6-month-old exclusively breast-fed infants is necessary, and information on the metabolisability of breast milk in mid-infancy is desirable. This evidence should inform future recommendations on infant feeding and help to identify research needs in infant energy balance.