The effect of long-term dietary protein excess and deficit on whole-body protein-N turnover (WBPNT) was examined in lambs nourished by intragastric infusions of nutrients. Ten sheep were given 500 mg N/kg metabolic weight (W0.75) per d from casein for 2 weeks and then either 50 (L), 500 (M) or 1500 (H) mg N/kgW0.75 per d for 6 weeks. Volatile fatty acids were infused at 500 kJ/kgW0.75 per d. Daily WBPNT was measured by continuous intravenous infusion of [l-13C]leucine 3 d before, and on days 2, 21 and 42 after the alteration in protein intake. Whole-body protein-N synthesis (WBPNS) was calculated as the difference between WBPNT and the protein-N losses as urinary NH3 and urea. Whole-body protein-N degradation (WBPNS) was then estimated from WBPNS minus protein gain determined from N balance. Fractional rates of WBPNS and WBPND were calculated against fleece-free body N content. WBPNS rates at the L, M and H intakes were respectively 35·1, 41·5 amd 6·37 g/d (P < 0.001) on average over the 6 weeks and WBPND rates were 39·5, 41·1 and 56·8 g/d (P < 0.001). The fractional rates of WBPNS were 5·01, 6·37 and 7·73% per d (P < 0.001) while those of WBPND were 5·64, 6·29 and 6·81% per d (P < 0.005) respectively. On days 2, 21 and 42, WBPNS rates at intake H were 54·0, 61·8 and 75·4 g/d (P = 0·03) respectively, and WBPND rates were 43·2, 56·4 and 70·9 g/d (P = 0.03); at intake L the amounts were 38·2, 34·2 and 32·8 g/d for WBPNS (P = 0.003) and for WBPND were 43·4, 38·0 and 36·9 g/d (P = 0·016) respectively. There were no significant (P > 0·05) differences in fractional rates of WBPNS and WBPND with time at either the L or H intake. We concluded that absolute protein turnover was affected both by dietary protein intake and by body condition while the fractional rate of turnover was predominantly influenced by intake.