Objective To compare three simple methods of body composition analysis for the assessment of body fat in patients on long-term hemodialysis therapy. Design Cross-sectional study using the skinfold thickness, bioelectrical impedance analysis, and near-infrared interactance techniques after a hemodialysis session. Subjects/setting Ninety clinically stable patients (57 male/33 female) undergoing hemodialysis at the Dialysis Unit of the Federal University of São Paulo. Statistical analysis Analysis of variance, intraclass correlation coefficient, and Bland-Altman plot analysis were used for the comparative analysis between the methods. Results Body fat measurements obtained by skinfold thickness (13.5±6.2 kg) and bioelectrical impedance analysis (13.7±6.7 kg) were similar, whereas those measured by near-infrared interactance (11.3±5.1 kg) were significantly lower in comparison with skinfold thickness and bioelectrical impedance analysis (P<.001). The strongest intraclass correlation coefficient was found between bioelectrical impedance analysis and skinfold thickness (r=0.87), and near-infrared interactance vs skinfold thickness and bioelectrical impedance analysis methods yielded r=0.78 and r=0.76, respectively. Near-infrared interactance showed a progressive underestimation of body fat values in comparison with the bioelectrical impedance analysis technique in patients with higher amount of adiposity. Conclusion In our study, we cannot consider that one method of body composition analysis is more accurate than the other because we did not apply a gold standard method. However, the most simple, long-established, and inexpensive method of skinfold thickness seems to be still very useful to the dietitians' routine for assessing body fat in patients on long-term hemodialysis therapy. J Am Diet Assoc. 2003;103:195-199.