A comparison was made between the dose-to-the-mother deuterium-dilution method and the conventional test-weighing technique for determining human-milk intake in five exclusively breast-fed infants and in four breast-fed infants who received supplemental foods. After administration of 2H to the mothers human milk and infant urine were sampled over 14 d and analyzed for 2H:1H ratios by gas-isotope-ratio mass spectrometry. Infant total body water was determined by 18O dilution. The test-weighing procedure was conducted for 5 d consecutively. The intake of human milk (mean +/- SD) estimated by 2H dilution was 648 +/- 63 g/d and estimated by test-weighing was 636 +/- 84 g/d. The mean difference between the two methods was not significantly different from 0. The 2H-dilution and test-weighing techniques provide similar estimates of human-milk intake.