It has been suggested that the periparturient breakdown of immunity to parasites has a nutritional basis. Our overall hypothesis is that it results from a prioritised scarce nutrient allocation to reproductive functions (e.g. milk production) rather than to immune functions. We tested this hypothesis by offering five levels of dietary metabolisable protein, ranging from 0.65 to 1.25 times their assumed requirements, for 4 weeks post-parturition to twin-rearing Greyface ewes, experimentally infected with Teladorsagia circumcincta. We hypothesised that the initial increments of metabolisable protein supply would increase milk production without affecting the degree of breakdown of immunity whilst later increments would reduce the degree of breakdown of immunity. The first two increments of metabolisable protein supply indeed increased milk production and did not affect final worm burdens, but in contrast to the expectation, reduced faecal egg counts and total egg output. The last two increments of metabolisable protein supply did not further affect milk production and egg output, but resulted in reduced final worm burdens. Metabolisable protein supply did not affect plasma IgG and IgE antibody against somatic L(3) antigen but the first three increments reduced plasma pepsinogen and plasma IgA antibody. The last increment did not further reduce plasma pepsinogen but increased plasma IgA. Metabolisable protein supply did not systematically affect abomasal mucosal mast cell, globule leukocyte and eosinophil counts. Our results support the view that the priority of scarce metabolisable protein allocation to milk production over immune functions may be gradual rather than absolute. The contrast between effects of metabolisable protein supply on faecal egg count and final worm burden points towards the possibility that if different effector responses regulate fecundity and worm expulsion, then they would differ in their sensitivity towards changes in the degree of nutrient scarcity.