We studied the effects of supplementing N as distillers dried grains with solubles (DDGS) or urea to steers consuming corn-based diets. Six ruminally and duodenally cannulated steers (244 kg) were used in 2 concurrent 3 x 3 Latin squares and fed 1 of 3 corn-based diets: control (10.2% CP), urea (13.3% CP), or DDGS (14.9% CP). Periods were 14 d, with 9 d for adaptation and 5 d for collection of urine and feces. Urinary (15)N(15)N-urea enrichments, resulting from venous infusions of (15)N(15)N-urea, were used to measure urea kinetics. Dry matter intake (6.0 kg/d) was not affected by treatment, but N intake differed (99, 151, and 123 g/d for the control, DDGS, and urea treatments, respectively). Urea-N synthesis tended to be greater (P = 0.09) for DDGS (118 g/d) than for the control treatment (52 g/d), with the urea treatment (86 g/d) being intermediate. Urea-N excreted in the urine was greater (P < 0.03) for the DDGS (35 g/d) and urea treatments (29 g/d) than for the control treatment (13 g/d). Gastrointestinal entry of urea-N was not statistically different among treatments (P = 0.25), but was numerically greatest for DDGS (83 g/d), intermediate for urea (57 g/d), and least for the control (39 g/d). The amount of urea-N returned to the ornithine cycle tended to be greater (P = 0.09) for the DDGS treatment (47 g/d) than for the urea (27 g/d) or control treatment (16 g/d). The fraction of recycled urea-N that was apparently used for anabolism tended (P = 0.14) to be greater for the control treatment (0.56) than for the DDGS treatment (0.31), with the urea treatment (0.45) being intermediate, but no differences were observed among treatments in the amount of urea-N used for anabolism (P = 0.66). Urea kinetics in cattle fed grain-based diets were largely related to the amount of N consumed. The percentage of urea production that was captured by ruminal bacteria was greater (P < 0.03) for the control treatment (42%) than for the DDGS (25%) or urea treatment (22%), but the percentage of duodenal microbial N flow that was derived from recycled urea-N tended (P = 0.10) to be greater for the DDGS treatment (35%) than for the urea (22%) or control treatment (17%). Thus, ruminal microbes were more dependent on N recycling when the protein supplement was largely resistant to ruminal degradation.