The objectives of this study were to assess the relationship between urinary urea N (UUN) excretion (g/d) and milk urea N (MUN; mg/dL) and to test whether the relationship was affected by stage of lactation and the dietary crude protein (CP) content. Twelve lactating multiparous Holstein cows were randomly selected and blocked into 3 groups of 4 cows intended to represent early [123 +/- 26 d in milk (DIM); mean +/- standard deviation], mid (175 +/- 3 DIM), and late (221 +/- 12 DIM) lactation stages. Cows within each stage of lactation were randomly assigned to a treatment sequence within a split-plot Latin square balanced for carryover effects. Stage of lactation formed the main plots (squares) and dietary CP levels (15, 17, 19, and 21% of diet dry matter) formed the subplots. Graded amounts of urea were added to the basal total mixed ration to linearly increase dietary CP content while maintaining similar concentrations of all other nutrients among treatments. The experimental periods lasted 7 d, with d 1 to 6 used for adjustment to diets and d 7 used for total collection of urine as well as milk and blood sample collection. Dry matter intake and yields of milk, fat, protein, and lactose declined progressively with lactation stage and were unaffected by dietary CP content. Milk and plasma urea-N as well as UUN concentration and excretion increased in response to dietary CP content. Milk and urine urea-N concentration rose at increasing and decreasing rates, respectively, as a function of plasma urea-N. The renal urea-N clearance rate differed among lactation stages and dietary CP contents. The relationship between UUN excretion and MUN differed among lactation stages and diverged from linearity for cows in early and late lactation. However, these differences were restricted to very high MUN concentrations. Milk urea N may be a useful tool to predict the UUN excretion and ultimately NH(3) emission from dairy cattle manure.