Crossbred steers (n = 140; 353 kg) were implanted as follows: (C) control, no implant; (S) Synovex-S (20 mg estradiol benzoate + 200 mg progesterone); (R) Revalor (20 mg estradiol + 140 mg trenbolone acetate [TBA]; international dosage); (ST) S + finaplix-S (140 mg TBA); (STT) S + finaplix-S with finaplix-S reimplanted on d 58. Steers were harvested after 119 to 126 d on feed. Left sides of 40 carcasses were fabricated into boneless subprimals for two s.c. fat levels: 2.5 and .64 cm. Steers administered an estrogen (estradiol benzoate or estradiol) plus TBA gained more rapidly (P < .05) than C or S steers. Feed efficiency was improved (P < .05) with an estrogen plus TBA. No differences (P > .05) were noted among treatments for carcass s.c. fat thickness, percentage internal fat, or lean color. Carcasses from steers receiving TBA had larger (P < .05) longissimus areas and tended to have lower (P < .10) marbling scores and yield grades than C or S steers. Steers implanted with R had a lower (P < .05) percentage of U. S. Choice carcasses (51.8%) than C, S, and ST steers (82 to 86%). Shear force values for implanted steers tended to be higher (P < .10) than for controls. Implants increased (P < .05) subprimal and total side lean yields (.64 cm) compared to controls; the largest increases of 2.3 and 2.8%, respectively, occurred in steers receiving TBA plus an estrogen. Estrogen plus TBA exhibited favorable effects on gain, efficiency, and composition; however, the single estradiol plus TBA implant (R) decreased quality grade.