Heifers were adapted to Pinpointer 4000B Feeders 4 prior to the part of the study so that feed could be measured individually. They were penned within treatment group in groups of 12 with one pen per treatment and were rotated among pens at 28-d intervals to reduce effects due to pinpointer machine and pen. Diets were 40% roughage for the first 28 d, 25% roughage for the second 28 d, and 17% roughage for the remainder of the study (Table 1). All the diets were completely mixed and balanced to meet or exceed NRC (1984) requirements for an ADG of 1.1 kg.At the conclusion of the study (d 122), the heifers were weighed full on two consecutive days. After 35 d with no implants, all heifers were slaughtered in a commercial processing facility. At 24 h postmortem, carcasses were evaluated for quality and yield grade traits (USDA, 1975) by University of Illinois per- sonnel. Data were evaluated using a completely randomized design with four treatments. The General Linear Model, least-squares means and probability of difference procedures of SAS (1982) were used. The model statement included treatment as the independent variable with the remainder used as the residual error term.