To study metabolic adaptation to harsh foraging conditions, an experiment was conducted to characterize and quantify N utilization efficiency and urea metabolism in Tibetan and fine-wool sheep fed 4 levels of dietary N (11.0, 16.7, 23.1, and 29.2 g N/kg DM) in 2 concurrent 4 × 4 Latin square designs. Urea kinetics were determined using continuous intrajugular infusions of 15N15N-urea. Urinary excretions of total N and urea N increased linearly (P < 0.001) with dietary N and were not different between breeds (P ≥ 0.37). Fecal N excretion increased with dietary N for Tibetan sheep but not for fine-wool sheep (linear dietary N × breed; P < 0.05). Nitrogen retention (both amount per day and percentage of N intake) increased with increasing dietary N concentration (P < 0.001), and the rates of increase were greater in fine-wool than in Tibetan sheep (linear dietary N × breed and cubic dietary N × breed; P < 0.05). In Tibetan sheep, N retention as a percentage of intake was greatest for diets containing 16.7 g N/kg DM, whereas it was maximal for fine-wool sheep when the diet contained 23.1 g N/kg DM. Urea N entry rate, urea N recycled to the gastrointestinal tract (GIT), and urea N returned to the ornithine cycle all increased with dietary N (P < 0.05), and all were greater in Tibetan than fine-wool sheep for the 11.0 g N/kg DM diet but were greater in fine-wool than Tibetan sheep for the diet with 29.2 g N/kg DM (linear dietary N × breed; P < 0.05). Urea N excreted in feces, both amount and fraction of GIT entry rate, was less in Tibetan than fine-wool sheep for the 11.0 and 16.7 g N/kg DM diets but similar for diets with 23.1 or 29.2 g N/kg DM (linear dietary N × breed; P < 0.01). For the lowest-protein diet, the fraction of urea N production recycled to the GIT was greater in the Tibetan than fine-wool sheep (88% vs. 82%), but for the diet with 29.2 g N/kg DM it was greater for fine-wool than Tibetan sheep (46% vs. 39%; linear dietary N × breed; P < 0.05). Plasma urea N increased more rapidly in response to increasing dietary N concentration for fine-wool sheep than for Tibetan sheep (linear dietary N × breed; P < 0.05). Urea tubular load and the amount and percentage of urea reabsorbed by the kidney were greater in Tibetan than fine-wool sheep (P < 0.05). These results suggest that Tibetan sheep have mechanisms that allow them to utilize N more efficiently than the fine-wool sheep when dietary N is inadequate.