Manifestations of infectious disease may represent host adaptations to avoid or reduce the effects of infection on host fitness, parasite manipulations that benefit the pathogen's fitness, or nonadaptive side effects of parasitism. Threespine stickleback fish (Gasterosteus aculeatus) from Alaska and the cestode macroparasite Schistocephalus solidus provide an excellent system for study of the effects of parasitism on host egg size because females in populations there are capable of producing clutches of eggs in the face of substantial infection, contrary to the inhibition of reproduction that has been observed in other stickleback populations or other species of fish. A side effect resulting in reduction of mean ovummass among infected females was predicted based on the egg production process in female stickleback, the considerable energy and resource demands of S. solidus, and the chronic and progressive nature of the effects the macroparasite should have on the host fish. In each of 9 populations of G. aculeatus representing replicate natural experiments in lakes scattered across the Matanuska-Susitna Valley and the Kenai Peninsula of south-central Alaska and among all populations combined, the mean ovum mass of infected female fish is significantly reduced in comparison with that of uninfected females taken from the same population at the same time. Reduction in mean female egg mass ranged from 8 to 32% across all populations. To examine whether reduction in mean female ovum mass was a nonadaptive side effect or an adaptation, relatively large data sets from 2 of the populations were used. Mean ovum mass of infected females was predicted to decrease directly in relation to parasite index (PI) if the diminution in mean egg mass were the result of a nonadaptive side effect resulting from host nutrient loss. Alternatively, the absence of a relationship between PI and reduction in ovum mass is predicted if decreases in mean female ovum mass result from host or parasite adaptation (or both) because lightly infected hosts should show a response similar to that of heavily infected ones. In each of the 2 populations, there is a significant, negative relationship between mean female ovum mass and PI, demonstrating a correlation between the decrease in ovum mass and the level of infection. Thus, the results are consistent with the hypothesis that the reductions in mean female egg mass represent side effects of parasitism involving nutrient theft. Moreover, the proportional decline in egg mass with increasing PI apparently differed between the 2 populations, and there was no significant relationship between mean percent decrease in mean female ovum mass and mean PI across populations. These observations suggest that unknown ecological and evolutionary factors influence the degree of reduction in mean ovum mass in a population-specific manner.