Artificial selection for improved productivity may reduce an animal's ability to cope with pathogens. Here, we used Roslin mice, uniquely divergently selected for high (ROH) and low (ROL) body weight, to assess interactive effects of differing growth potential and protein nutrition on host resilience and resistance. In a 2 x 2 x 6 factorial design, ROH and ROL mice were either sham-infected or infected with 250 L(3)Heligmosomoides bakeri and fed diets with 30, 80, 130, 180, 230 and 280 g crude protein per kg. The infected ROL-30 treatment resulted in clinical disease and was discontinued. In the remaining ROL mice, infection and feeding treatments did not affect growth but infection reduced weight gain in ROH-30, ROH-80 and ROH-130 mice. Although infection resulted in temporarily reduced food intake (anorexia) in both mouse lines, mean food intake over the whole experiment was reduced in ROH mice only. ROH mice excreted more worm eggs and had higher worm burdens, with relatively fewer female worms, than ROL mice. However, these resistance traits were not sensitive to dietary protein. These results support the view that selection for high growth may reduce the ability to cope with pathogens, and that improved protein nutrition may to some extent ameliorate this penalty.