Updates have been made to the Cornell Net Carbohydrate and Protein System which includes estimations of recycled urea nitrogen, rumen protozoa production and yield, endogenous N secretions and a new system for calculating post-ruminal N digestion. The objective was to evaluate the ability of the updated model to balance diets of high producing dairy cattle below or close to requirements for both rumen N and EAA. Sixty-four high producing dairy cows (100 ± 31 DIM) were randomly assigned to one of 4 treatments: 1) limited in Met, MP and rumen N (Base); 2) adequate in Met but limited MP and rumen N (Base+M); 3) adequate in Met and rumen N, but limited MP (Base+MU); 4) adequate in MP, rumen N and balanced for all EAA (Positive). Dietary CP was 13.5, 13.6, 14.6 and 15.6% DM and model predicted dietary MP balance was −231, −310, −142 and 33 g/d for the Base, Base+M, Base+MU and Positive treatments, respectively. No differences were observed in DMI or milk yield (24.1 to 24.8 and 40.0 to 41.8 kg/d, respectively). Energy corrected milk, fat and true protein yield were greater (3.3, 0.09 and 0.11 kg/d, respectively; P < 0.001) in cows fed the Positive compared with the Base treatment. True protein concentration in milk was higher (P < 0.001) and milk fat tended to be higher (P < 0.10) in cows fed the Positive and Base+MU treatments than cows fed the Base and Base+M treatments. Bacterial growth was predicted to be depressed by 16% and 17% for the Base and Base+M treatments, respectively, due to the predicted rumen N balance and this prediction corresponded with lower (P < 0.05) apparent total-tract NDF digestion. The study demonstrated high levels of milk yield can be achieved when diets are formulated on rumen N balance and EAA, even when CP is less than 14% DM provided adequate AA are supplied to the small intestine. The highest milk yield was achieved when considering all EAA in the ration formulation process.