To determine the contribution of tissue protein reserves to lactational performance, multiparous female Sprague-Dawley rats were mated, caged individually and offered a diet high in protein (215 g crude protein (N × 6·25; CP)/kg dry matter (DM);H) ad lib. until day 12 of gestation. Subsequently half the rats continued to receive diet H while the remainder were offered a diet low in protein (65 g CP/kg DM;L) until parturition. This treatment aimed to produce a difference in carcass protein at parturition. On day 1 of lactation females were allocated to either diet H or a low-protein diet (90 g CP/kg DM; L2) offered until day 13 of lactation, giving four lactation treatment groups HH, HL2, LH and LL2. Groups of females were slaughtered on days 2 and 12 of gestation and days 1 and 13 of lactation and carcass and major organs were analysed. Weight gain of standardized litters was used as an indicator of lactational performance. Maternal carcass protein contents at parturition were 43·5 (SE 1·2) and 38·7 (SE 0·8) g (P < 0·01) for diets H and L respectively. During lactation there was little change in carcass protein content of HH rats while LH rats appeared to replenish their depleted reserves. Food intake or lactational performance did not differ between these two groups. HL2 and LL2 rats lost carcass protein with HL2 rats losing more than LL2 rats (P < 0·05). Intake and lactational performance were reduced compared with that on diet H (P < 0.05) but for the first 6 d of lactation were both greater (P < 0·05) for diet HL2 than for diet LL2. All four groups showed a considerable loss of body fat during lactation which was not affected by diet. The ability of HL2 rats to catabolize more protein and consume more food allowed them to sustain a greater lactationai performance. Previous maternal protein depletion had no influence on lactationai performance as long as an adequate supply of dietary protein was provided.