Data on habitat selection, overwinter survival, and litter survival rate of muskrats indicate that fitness differs among habitats in marsh environments. We measured nine nutrient-dependent phenotypic variables of muskrats and one independent assessment of forage quality during the ice-free period (May-October) from 1992 to 1994. Data were used to test the hypothesis that differences in food resources limit the demographic performance of muskrats in marginal versus prime habitats (food limitation hypothesis, FLH). None of the null predictions relating habitat quality with reproductive parameters could be rejected statistically. Adult body mass and length were greater in prime habitats, but fat content did not differ, which suggested that the difference was associated with larger structural size of dominant animals in prime habitats. Growth rate of weaned juveniles was not different among habitats, which represented the strongest evidence for rejecting the FLH. Faecal crude protein content, which was used as an index of quality of food ingested, was marginally different between the two most suitable habitats, but the direction of the difference did not support our prediction. Overall, the temporal variation in faecal crude protein content was similar among habitats. We conclude that food is not a key factor limiting the demographic performance of muskrats among habitats during the ice-free season.