To determine the validity and reliability of percent body fat estimates in 177 boys and 154 girls between 12-17 years of age, percent body fat was assessed once using dual-energy X-ray absorptiometry and twice using the sum of two skinfolds and three bioelectrical impedance analysis devices. The assessments were repeated on 79 participants on a second day. The agreement between the percent body fat estimates from the four prediction methods and dual-energy X-ray absorptiometry was evaluated using Bland-Altman analyses and a mixed linear model. All methods were reliable within and between days. The sum of two skinfolds (plus or minus 6.8% body fat), OMRON (OMRON Healthcare Inc., Vernon Hills, Illinois, USA) bioelectrical impedance analysis (0 plus or minus 7.3% body fat), and TANITA 521 (TANITA Corporation, Tokyo, Japan) bioelectrical impedance analysis (plus or minus 7.6% body fat) had wide prediction intervals and are acceptable for use in large population-based studies. The TANITA 300A bioelectrical impedance analysis is not recommended for use in children and adolescents due to its wide prediction interval (plus or minus 8.1% body fat) and large bias. A criterion method should be used when the accurate assessment of body composition of an individual is critical. (Contains 3 figures and 5 tables.)