The influence of energy source (silage- or grain-based) on empty body and carcass composition and adipocyte cellularity independent of rate of gain was tested. Sixty-four Angus steers were allotted to either a forage (ad libitum) or grain (limit-fed) diet for a growing phase (145 d) followed by 45, 75, or 105 d of ad libitum access to a grain-based diet. Eight steers were slaughtered initially and eight from each treatment were slaughtered at the end of the growing phase, and at each of the termination dates. The silage growing diet consisted (DM basis) of 55% sorghum silage (approximately 24% dry matter), 22% alfalfa hay, 11% ground shelled corn, and 11% soybean meal. The grain-based growing diet was composed of 77% ground shelled corn, 5% soybean meal, 14% cottonseed hulls, 3% molasses, and 1% salt and mineral; it was limit-fed to produce the same rate of gain as the silage diet. No implants or ionophores were used. At the end of the growing phase, the steers fed grain were heavier and had a higher percentage of fat in the empty body (24 vs 19% fat) and the carcass (26 vs 21% fat) than did steers fed forage. Rate of gain during the growth phase was related positively to percentage of carcass fat; when corrected for fill, data for both diets fit one regression line for fat vs rate of gain. When adjusted for gain during the growing phase, fat content was not different in empty body or carcass, but internal fat was higher (P < .10) for steers fed grain. After 45 d on the finishing diet, carcass fat remained low (23%), but after 75 and 105 d, fat content reached 27%. Source of energy did not detectably affect carcass composition independent of rate of gain. Cell size of adipocytes from four adipose depots increased with time on feed but were not affected by diet during the growing phase. Lean Choice beef can be produced in only 45 d in the feedlot with medium-framed Angus cattle.