In Reply.—We agree with Abrams et al that correction of BMC for BW may be useful in evaluating bone mineralization, especially for correcting cross-sectional data for differences in BW among infants of the same age. The usefulness of dividing by BW is evident in the data reported by Abrams et al. The cross-sectional population variance for radius BMC/BW was less than that for BMC, which suggests that BMC/BW might allow the detection of statistically significant differences between smaller groups of infants than would be possible using BMC. This advantage disappears, however, if the variance for BW is high, as in our studies of humerus BMC using the Norland 278A densitometer. The relatively poor reproducibility of BW measurements in our study prevented any potential improvement that might be gained by using BMC/BW instead of BW as an assessment of bone mineral status. The reproducibility of BW data may have been better in