Background/Objectives: To investigate the relationships between body mass index (BMI), body fatness, ethnicity, age and blood lipids. Subjects/Method: In a subsample (N ¼ 643) of the 2002 New Zealand Child Nutrition Survey (N ¼ 3275, age 5-14 years) consisting Māori (89 boys, 69 girls), Pacific (154 boys, 194 girls) and European (71 boys, 66 girls) ethnic groups, fat-free mass (FFM) was determined by bioimpedance analysis. FFM index (FFMI, FFM/height 2 ), fat mass index (FMI, fat mass/height 2 ) and percentage body fat (%BF) were derived. Plasma total cholesterol and high-density lipoprotein (HDL) cholesterol were measured in 515 (80%) of these children. Results: For the same BMI, %BF was lower in Pacific Island (Po0.0001) and Māori (Po0.0001) girls compared with European but for boys there were no ethnic differences. Cross-sectional analysis across ages using Hattori charts clearly showed sexual dimorphism with girls having a higher FMI, FFMI and BMI than boys by the age of 13 year. Both Pacific girls and boys showed upward trends in FFMI and FMI with age that were significantly higher than those of the other ethnicities. Adjusting for age, sex and ethnicity, %BF was a better predictor (R 2 ¼ 0.090, P ¼ 0.0002) of total cholesterol than BMI (R 2 ¼ 0.061, P ¼ 0.049) or waist (R 2 ¼ 0.075, P ¼ 0.013) while FFMI was the best predictor of HDL cholesterol (R 2 ¼ 0.15, Po0.0001) and waist was the best predictor of the total cholesterol/HDL cholesterol ratio (R 2 ¼ 0.17, Po0.0001). Conclusions: Measurements of body fat and indices of body fatness additional to BMI in multiethnic paediatric populations allow changes in fat mass and FFM to be tracked and improve the ability to predict dyslipidaemia.