True fractional Ca absorption (TFCA) was compared in children with different habitual Ca intakes using a double-label stable-isotope technique. Chinese children aged 7 years from Hongkong (n 22) and Jiangmen (n 12) participated in the study. An oral administration of 8 mg 44Ca in 100 g chocolate milk was given shortly after an intravenous injection of 0·75 mg 42Ca. Ca isotopic ratios were determined in urine samples collected 24 h later using thermal-ionization mass spectrometry. There was no significant difference in TFCA between Jiangmen and Hongkong children (P=0·16). TFCA of a lower-Ca-intake group (Ca > 500 mg/d, n 19) with mean Ca intake 359 mg/d was 63·1 (SD 10·7)%; and that of a higher-Ca-intake group (Ca > 500 mg/d, n 15) with mean Ca intake 862 mg/d was 54·8 (SD 7·3)%; the difference in TFCA was significant (P=0·016). Serum levels of 25-hydroxycholecalciferol of the children were adequate (33·7 (SD 7·7) ng/ml). The present study indicates that growing children accustomed to a low-Ca diet appear to be able to enhance their absorptive capacity. If it is assumed that dietary Ca absorption by Chinese children resembles their TFCA from a single meal of chocolate milk, then the recommended dietary allowance (RDA) for Ca for Chinese children would be lower than the US RDA (800 mg/d), which is based on an estimated 40% Ca absorption as reported for Caucasian children. A comparative absorption study is necessary to determine whether there is any difference in TFCA between Caucasian and Chinese children.