The relative merits of test weighing, water turnover methods, and a flowmeter method for the measurement of milk intake in breast-fed babies are reviewed to allow the prospective investigator to choose the method most suited to his or her needs. Provided that measurements are made over 3–4 days to minimize the effects of day-to-day variation in milk intake, test weighing is a satisfactory procedure when feed frequency is low and individual feed volumes are large. However, in developing countries where frequency is high and feed volumes low, test weighing is inherently less accurate and may impose an unfamiliar and unphysiological discipline on the mother and child that severely limits its usefulness. In these circumstances methods based on the measurement of water turnover rates using 2H2O are the only procedures likely to yield useful information. A method in which single doses of 2H2O are given to the mother, and milk intake rates measured over 14 days, is described. Neither test weighing nor water turnover methods provide simultaneous milk intake and composition data. The development of flowmeter methods will make this possible, but their use is likely to be limited to metabolic wards rather than the home and widespread use in community studies is not a practical proposition.