Abstract Twenty-four multiparous dairy cows (eight with ruminal cannulae) were blocked by days in milk and assigned to six balanced 4 x 4 Latin squares with 21-d periods. The four diets, formulated from alfalfa silage plus a concentrate mix based on ground high moisture ear corn, contained (dry matter basis): 1) 20% concentrate, 80% alfalfa silage (24% nonfiber carbohydrate; NFC), 2) 35% concentrate, 65% alfalfa silage (30% NFC), 3) 50% concentrate, 50% alfalfa silage (37% NFC), or 4) 65% concentrate, 35% alfalfa silage (43% NFC). Soybean meal and urea were added to make diets isonitrogenous with equal nonprotein nitrogen (NPN) (43% of total N). Total urine was collected with indwelling Folley catheters for 24h during each period. There was no effect of diet on urinary creatinine excretion (average 29 mg/kg of BW/d). There were quadratic effects of diet on total urinary ecretion of allantoin, uric acid, and purine derivatives (allantoin plus uric acid), and on ruminal synthesis of microbial N estimated from purine derivatives; maxima occurred at about 35% dietary NFC. Urinary excretion also was estimated with spot urine samples from creatinine concentration and the mean daily creatinine excretion. Daily excretion of allantoin, uric acid, and purine derivatives estimated from spot urine sampling followed the same pattern as that observed with total collection; differences between measured and estimated urine volume were significant only for 35% dietary concentrate. Spot urine sampling appeared to yield satisfactory estimates of purine derivative excretion. Maximal urea N excretion was estimated to occur at about 31% dietary NFC. Milk allantoin secretion increased linearly with concentrate and accounted for 4 to 6% of the total purine derivative excretion. Microbial yield was maximal at 35% dietary NFC, suggesting that this was the optimal level for utilization of dietary NPN from alfalfa silage and other sources.