This lysimeter experiment was designed to investigate the effects of dietary crude protein (CP) concentration on nitrate-N (NO-N) and ammonia (NH) losses from dairy manure applied to soil and manure N used for plant growth. Lactating dairy cows were fed diets with 16.7% CP (HighCP) or 14.8% CP (LowCP) content. Feces and urine were labeled with N by ruminal pulse-doses of NHCl. Unlabeled and N-labeled feces and urine were used to produce manure for a study with 21 lysimeters in a greenhouse. Manure application rate was 277 kg N ha. Ammonia emissions were measured at 3, 8, 23, 28, 54, and 100 h after manure application. Manure was incorporated into the soil, and a leaching event was simulated. Spring barley was planted (387 plants per m) 7 d after the leaching event and harvested at senescence. Ammonia emission rates and the contribution of urinary N to NO-N were on average about 100% greater for HighCP vs. LowCP manures. With both LowCP and HighCP manures, a greater proportion of urinary vs. fecal N was recovered in leachate NO-N. There was no difference in whole-crop barley N yields between LowCP and HighCP manures, but barley kernel N yield tended to be greater ( = 0.09) for lysimeters treated with HighCP manures. Using a unique labeling approach, this lysimeter experiment demonstrated that when applied at equal soil N application rates, manure from cows fed the HighCP diet resulted in markedly greater NH emissions and urinary N losses with leachate NO-N than manure from cows fed the LowCP diet.