PURPOSE: The body adiposity index (BAI) is a recent anthropometric measure proven to be valid in predicting body fat percentage (BF%) in some populations. However, the results have been inconsistent across populations. The study was designed to verify the validity of BAI in predicting BF% in a sample of obese adults, using dual-energy X-ray absorptiometry (DEXA) as the reference method. METHODS: A cross-sectional study was conducted in 48 participants (54% female, mean age 41.0±7.3 years old). DEXA was used as the “gold standard” to determine BF%. Pearson’s correlation coefficient was used to evaluate the association between BAI and BF%, as assessed by DEXA. A paired sample t-test was used to test differences in mean BF% obtained with BAI and DEXA methods. To evaluate the concordance between BF% as measured by DEXA and as estimated by BAI, we used Lin’s concordance correlation coefficient and Bland-Altman agreement analysis. RESULTS: The correlation between BF% obtained by DEXA and that estimated by BAI was r=0.844, p < 0.001. Paired t-test showed a significant mean difference in BF% between methods (BAI = 33.3±6.2% vs. DEXA 39.0±6.1; p < 0.001). The bias of the BAI was -6.0±3.0 BF% (95% CI= -12.0 to 1.0), indicating that the BAI method significantly underestimated the BF% compared to the reference method. Lin’s concordance correlation coefficient was considered stronger (ρc=0.923, 95% CI = 0.862 to 0.957). CONCLUSIONS: In obese adults BAI presented low agreement with BF% measured by DEXA; therefore, BAI is not recommended for BF% prediction in this Latin American obesity sample studied. Supported by Universidad del Rosario (Code N° FIUR DN-BG001)