One hundred and eight Holstein dairy heifers were divided into two weight blocks based on beginning body weight (E = < 182 kg, light, > 270 kg, heavy). Heifers within weight blocks were randomly assigned to one of five treatments. Treatments consisted of three levels of NDF (35%, 45%, 55%) at 100% ofNRC TDN recommendation, and three levels of TDN (85%, 100%, 115 of NRC) at 45% NDF for light heifers. For heavy heifers, treatments consisted of three levels of NDF (40%, 50%, 60%) at 100% NRC TDN, and three treatments with similar levels of TDN at 50% NDF. Heifers were fed total mixed rations of corn silage, alfalfa haylage, ground orchardgrass hay, soybean meal, high moisture corn and a mineral mix formulated for 0.68 kg ADG. Mean gains (kg/d) were 1.07 and .96 for the light and heavy blocks, respectively. DMI as a percent of BWT differed across NDF levels for light heifers (3r'= 2.96, 2.56, 2.57, 2.57, and 2.87 % of BWT for the low, med., high NDF, and low, high TDN treatments, respectively). Heavy heifer DMI% did not differ (.Â¥= 2.45, 2.40, 2.24, 2.15, and 2.42 % ofBWT for low, med., and high NDF, and low, high TDN treatments, respectively). Equations to predict DMI were developed using intake information from this and a previous study. Backward stepwise regression was utilized to generate a simplified model. Model selected was: DMI (kg/d) = -5.9781 + (2.2120E-05 * BWTSQ) - (5.5527 * GAIN) + (2.7837 * GAINSQ) + (0.4668 * NDF) + (5.3930 * NDFSQ) + (0.03285 * DM) + 1 (7.7859E-03 * BWT * GAIN);11 = 514, rz = .67.