It is expensive and laborious to evaluate carcass composition in beef cattle. The objective of this study was to evaluate a method to predict the 9th to 11th rib section (rib9–11) composition through empirical equations using dual energy X-ray absorptiometry (DXA). Dual energy X-ray absorptiometry is a validated method used to describe tissue composition in humans and other animals, but few studies have evaluated this technique in beef cattle, and especially in the Zebu genotype. A total of 116 rib9–11 were used to evaluate published prediction equations for rib9–11 composition and to develop new regression models using a cross-validation procedure. For the proposed models, 93 ribs were randomly selected to calculate the new regression equations, and 23 different ribs were randomly selected to validate the regressions. The rib9–11 from left carcasses were taken from Nellore and Nellore × Angus bulls from 3 different studies and scanned using DXA equipment (GE Healthcare, Madison, WI) in the Health Division at Universidade Federal de Viçosa (Viçosa, Brazil). The outputs of the DXA report were DXA lean (g), DXA fat free mass (g), DXA fat mass (g), and DXA bone mineral content (BMC; g). After being scanned, the rib9–11 were dissected, ground, and chemically analyzed for total ether extract (EE), CP, water, and ash content. The predictions of rib fat and protein from previous published equations were different (P < 0.01) from the observed composition. New equations were established through leave-one-out cross-validation using the REG procedure in SAS. The equations were as follows: lean (g) = 37.082 + 0.907× DXA lean (R2 = 0.95); fat free mass (g) = 103.224 + 0.869 × DXA fat free mass (R2 = 0.93); EE mass (g) = 122.404 + 1.119 × DXA fat mass (R2 = 0.86); and ash mass (g) = 18.722 + 1.016 × DXA BMC (R2 = 0.39). The equations were validated using Mayer's test, the concordance correlation coefficient, and the mean square error of prediction for decomposition. For both equations, Mayer's test indicated that if the intercept and the slope were equal to 0 and 1 (P > 0.05), respectively, then the equation correctly estimated the rib composition. Comparing observed and predicted values using the new equations, Mayer's test was not significant for lean mass (P = 0.26), fat free mass (P = 0.67), EE mass (P = 0.054), and ash mass (P = 0.14). We concluded that the rib9–11 composition of Nellore and Nellore × Angus bulls can be estimated from DXA using the proposed equations.