Carcass measurements from 1,664 steers from the Germ Plasm Utilization project at U.S. Meat Animal Research Center were used to estimate heritabilities (h(2)) of, and genetic correlations (r(g)) among, 14 carcass traits adjusted to different endpoints (age, carcass weight, and fat thickness): HCW (kg), dressing percent (DP), adjusted fat thickness (AFT, cm), LM area (LMA, cm(2)), KPH (%), marbling score (MS), yield grade (YG), predicted percentage of retail product (PRP), retail product weight (RPW, kg), fat weight (FW, kg), bone weight (BNW, kg), actual percentage retail product (RPP), fat percent (FP), and bone percent. Fixed effects in the model included breed group, feed energy level, dam age, birth year, significant (P < 0.05) interactions, covariate for days on feed, and the appropriate covariate for endpoint nested (except age) within breed group. Random effects in the model were additive genetic effect of animal and total maternal effect of dam. Parameters were estimated by REML. For some traits, estimates of h(2) and phenotypic variance changed with different endpoints. Estimates of h(2) for HCW, DP, RPW, and BNW at constant age, weight, or fat thickness were 0.27, -, and 0.41; 0.19, 0.26, and 0.18; 0.42, 0.32, and 0.50; and 0.43, 0.32, and 0.48, respectively. Magnitude and/or sign of r(g) also changed across endpoints for 54 of the 91 trait pairs. Estimates for HCW-LMA, AFT-RPW, LMA-YG, LMA-PRP, LMA-FW, LMA-RPP, and LMA-FP at constant age, weight, or fat thickness were 0.32, -, and 0.51; -0.26, -0.77, and -; -0.71, -0.89, and -0.66; 0.68, 0.85, and 0.63; -0.16, -0.51, and 0.22; 0.47, 0.57, and 0.27; and -0.44, -0.43, and -0.18, respectively. Fat thickness was highly correlated with YG (0.86 and 0.85 for common age and weight) and PRP (-0.85 and -0.82 for common age and weight), indicating that selection for decreased fat thickness would improve YG and PRP. Carcass quality, however, would be affected negatively because of moderate r(g) (0.34 and 0.35 for common age and weight) between MS and AFT. Estimates of h(2) and phenotypic variance indicate that enough genetic variation exists to change measures of carcass merit by direct selection. For some carcass traits, however, magnitude of change would depend on effect of endpoint on h(2) and phenotypic variance. Correlated responses to selection would differ depending on endpoint.