1. Captive propagation can lead to phenotypic change in fish populations, but the broader community-level consequences of captive phenotypes remain largely unknown. 2. Here we investigate the degree to which captive propagation alters the phenotypes and ecological roles of fish stocked into wild communities. We focus on captive propagation of western mosquitofish (Gambusia affinis) for biocontrol, which represents one of the largest-scale production efforts for any fish released into the wild. 3. Captive propagation in mosquitofish consistently generated novel mixtures of morphological and behavioral traits that deviate from those of wild populations. 4. A mesocosm experiment showed that mosquitofish from captive propagation facilities differ from wild fish in their effects on aquatic community structure by shifting their consumption to less-mobile, benthic prey. 5. Synthesis and applications: These findings provide evidence that captive-propagated and translocated wild stocks not only differ in phenotype, but can have substantially different ecological effects on the communities into which they are introduced. Therefore, captive propagation programs involving continual release should expand their concerns beyond altered phenotypes and fitness to include whether propagated fish actually provide the intended ecological roles and services associated with wild counterparts. Infusions of wild alleles and captive environments that mimic wild conditions are recommended strategies to retain the desired ecological role of captive propagated fish