Silva, AM, Fields, DA, Quitério, AL, and Sardinha, LB. Are skinfold-based models accurate and suitable for assessing changes in body composition in highly trained athletes? J Strength Cond Res 23(6): 1688-1696, 2009-This study was designed to assess the usefulness of skinfold (SKF) equations developed by Jackson and Pollock (JP) and by Evans (Ev) in tracking body composition changes (relative fat mass [%FM], absolute fat mass [FM], and fat-free mass [FFM]) of elite male judo athletes before a competition using a 4-compartment (4C) model as the reference method. A total of 18 male, top-level (age: 22.6 ± 2.9 yr) athletes were evaluated at baseline (weight: 73.4 ± 7.9 kg; %FM4C: 7.0 ± 3.3%; FM4C: 5.1 ± 2.6 kg; and FFM4C: 68.3 ± 7.3 kg) and before a competition (weight: 72.7 ± 7.5 kg; %FM4C: 6.5 ± 3.4%; FM4C: 4.8 ± 2.6 kg; and FFM4C: 67.9 ± 7.1 kg). Measures of body density assessed by air displacement plethysmography, bone mineral content by dual energy X-ray absorptiometry, and total-body water by bioelectrical impedance spectroscopy were used to estimate 4C model %FM, FM, and FFM. Seven SKF site models using both JP and Ev were used to estimate %FM, FM, and FFM along with the simplified Ev3SKF site. Changes in %FM, FM, and FFM were not significantly different from the 4C model. The regression model for the SKF in question and the reference method did not differ from the line of identity in estimating changes in %FM, FM, and FFM. The limits of agreement were similar, ranging from −3.4 to 3.6 for %FM, −2.7 to 2.5 kg for FM, and −2.5 to 2.7 kg for FFM. Considering the similar performance of both 7SKF- and 3SKF-based equations compared with the criterion method, these data indicate that either the 7- or 3-site SFK models are not valid to detect %FM, FM, and FFM changes of highly trained athletes. These results highlighted the inaccuracy of anthropometric models in tracking desired changes in body composition of elite male judo athletes before a competition.