Abstract Most pathogens are generalists capable of infecting multiple host species or strains. Trade‐offs in performance among different hosts are expected to limit the evolution of generalism. Despite the commonness of generalism, the variation in infectivity, transmission, and trade‐offs in performance among host species have rarely been studied in the wild. To understand the ecological and evolutionary drivers of multi‐host pathogen infectivity and transmission potential, I studied disease severity, transmission dynamics, and infectivity variation of downy mildew pathogen Peronospora sparsa on its three host plants Rubus arcticus , R. chamaemorus , and R. saxatilis . In a survey of 20 wild and cultivated sites of the three host species, disease severity varied by host species and by host population size but not among wild and cultivated sites. To understand how alternative host presence and plant diversity affect transmission of the pathogen, I conducted a transmission experiment. In this experiment, alternative host abundance and plant diversity together modified P. sparsa transmission to trap plants. To understand how resistance to P. sparsa varies among host species and genotypes, I conducted an inoculation experiment using 10 P. sparsa strains from different locations and 20 genotypes of the three host species. Significant variation in infectivity was found among host genotypes but not among host species. When trade‐offs for infectivity were tested, high infectivity in one host species correlated with high infectivity in another host species. However, when pathogen transmission‐related life‐history correlations were tested, a positive correlation was found in R. arcticus but not in R. saxatilis . The results suggest that host resistance may shape pathogen life‐history evolution with epidemiological consequences in a multi‐host pathogen.