The Cornell Net Carbohydrate and Protein System (CNCPS) prediction of fiber digestion and microbial mass production from ruminally degraded carbohydrate has been adjusted to accommodate a ruminal N deficiency. The steps for the adjustment are as follows: 1) the ruminal available peptide and ammonia pools are used to determine the N allowable microbial growth; 2) this value is subtracted from the energy allowable microbial growth to obtain the reduction in microbial mass; 3) this mass reduction is allocated between pools of bacteria digesting fiber (FC) and nonfiber (NFC) carbohydrate according to their original proportions in the energy allowable microbial growth; 4) the reduction in fermented FC is computed as the FC bacterial mass reduction divided by its yield (g bacteria/g FC digested); and 5) this reduction is added to the FC fraction escaping the rumen. Five published studies included information that allowed us to evaluate the response of animals to added dietary N. These evaluations compared observed and CNCPS-predicted ADG with and without this adjustment. The adjustment decreased the CNCPS overprediction of ADG from 19.2 to 4.7%, mean bias declined from .16 to .04 kg/d, and the r2 of the regression between observed and metabolizable energy (ME) or metabolizable protein allowable ADG was increased from .83 to .88 with the adjustment. When the observed dry matter intake was regressed against CNCPS-predicted DMI with an adjustment for reduction in cell wall digestibility, the r2 was increased from .77 to .88. These results indicated the adjustment for ruminal nitrogen deficiency increased the accuracy of the CNCPS model in evaluating diets of growing animals when ruminally degraded N is deficient.