Test-weigh measurements of 24 h milk intake by babies in the immediate puerperium (days 1-9) were carried out in a group of 18 mothers and babies using an accurate electronic balance. Estimated milk intakes were then calculated on the basis of one, and of two consecutive, breast feeds. These estimates were then compared with the actual total intake to give a measurement of reliability. Significant positive correlations were found between both the one-feed and the two-feed estimated intake, and the measured 24 h intake. Accuracy increased from day 3 onwards. These results show that using an electronic balance an accurate estimate of 24 h milk intake during the first week of life can be obtained by using one or two test-weigh measurements. While this gives a minimally invasive and reliable method for estimating daily milk intake, it is not advocated as a routine procedure in postnatal wards.