Infant test weighing and maternal test weighing are two independent methods for determining milk intake by the breast-fed infant. The sources of error in both these test weighing methods were examined with particular emphasis on the importance of evaluating and correcting for evaporative water loss (EWL). EWL ranged from 3 to 94% of the mother's change in weight after a single breast feed and from 3 to 55% of the infant's change in weight after a single breast feed. Correcting for EWL during a breast feed involved determining the time between the pre- and postfed weighings and measuring the rate of EWL after breast feeding. Significant correlations (p less than 0.001) were found between milk volume intake measured by test weighing the mother and correcting for maternal EWL, and milk volume intake measured by test weighing the infant and correcting for infant EWL. An improved method for measuring 24-h milk intakes by maternal test weighing using a sensitive electronic balance and correcting for EWL is described. The milk intakes, corrected for EWL, ranged from 690-1,041 g/24 h. If no correction for EWL was made then the average overestimate of milk intake by maternal test weighing was 14 +/- 6%.