Eight multicatheterized wethers (35.9 +/- .8 kg BW) were used in a replicated 4 x 4 Latin square design to measure N retention and net uptake and release of plasma metabolites across the portal-drained viscera (PDV), hepatic (HEP), and total splanchnic (TS) tissues in response to changes in supplemental N source. Treatments selected to provide different amounts of undegradable intake protein (UIP) were urea, soybean meal (SBM), poultry by-product meal (PBM), and bloodmeal:corn gluten meal (BMCGM; 50:50 CP basis). Diets (urea, SBM, PBM, and BMCGM) contained 12.9, 13.8, 13.6, and 13.2% CP, respectively. Periods were 10 d, with total feces and urine collected on d 7 to 10 and blood sampled on d 10. Wethers were fed at 2% of BW in 12 daily portions. Nitrogen retention was 2.2, 3.3, 4.1, and 4.4 g/d for urea, SBM, PBM, and BMCGM, respectively. Urea had less (P < .01) N retention than SBM, PBM, and BMCGM; SBM had less N retention (P < .01) than PBM and BMCGM. Arterial, portal, and hepatic plasma flows were greater (P < .09) for SBM than for PBM and BMCGM (21 vs 16, 17; 84 vs 72, 72; 105 vs 87, 88 L/h). Portal plasma flow was greater (P < .10) for urea than for SBM, PBM, and BMCGM (85 vs 84, 72, 72 L/h). Portal-drained viscera and TS alpha-amino N (AAN) fluxes were less (P < .05) for PBM than for BMCGM (20.5 vs 26.6 and 7.2 vs 15.1 mmol/h), but TS AAN flux was less (P < .05) for urea than for SBM, PBM, and BMCGM (6.9 vs 16.9, 7.2, 15.1 mmol/h). Portal-drained viscera flux and HEP removal of NH3 N were greater (P < .001) for SBM than for PBM and BMCGM (27.7 vs 19.4, 20.6; -28.1 vs -20.0, -21.4 mmol/h). Gut use was less (P = .07) and HEP and TS fluxes of urea N were greater (P < .01) for SBM than for PBM and BMCGM (-4.92 vs -8.32, -7.93; 25.87 vs 16.54, 20.00; 20.95 vs 8.22, 12.07 mmol/h). These data suggest that PBM and BMCGM improved efficiency of N use compared with urea and SBM by reducing urinary N loss.