It has been suggested that the harm parasites cause to their hosts is an unavoidable consequence of parasite reproduction with costs not only for the host but also for the parasite. Castrating parasites are thought to minimize their costs by reducing host fecundity, which may minimize the chances of killing both host and parasite prematurely. We conducted a series of experiments to understand the evolution of virulence of a castrating bacterium in the planktonic crustacean Daphnia magna. By manipulating food levels during the infection of D. magna with the bacterium Pasteuria ramosa, we showed that both antagonists are resource-limited and that a negative correlation between host and parasite reproduction exists, indicating resource competition among the antagonists. Pasteuria ramosa also induces enhanced growth of its hosts (gigantism), which we found to be negatively correlated with host fecundity but positively correlated with parasite reproduction. Because infected hosts never recovered from infections, we concluded that gigantism is beneficial only for the parasite. Hosts, however, have evolved counteradaptations. We showed that infected hosts have enhanced reproduction before castration. This shift to earlier reproduction increases overall host fecundity and compromises parasite reproduction. Finally, we showed that this resource conflict is subject to genetic variation among host and parasite genotypes within a population and is therefore likely to be an important force in the coevolution of virulence in this system. A verbal model is presented and suggests that the adaptive value of gigantism is to store host resources, which are liberated after parasitic castration for later use by the growing parasite. This hypothesis assumes that infections are long lasting, that is, that they have a high life expectancy.