Abstracta Classical biological control programs introduce primary parasitoids into new geographic regions, often exposing them to existing populations of hyperparasitoids. Hyperparasitoids are frequently implicated in the failure of parasitoid biological control agents to establish and provide control of insect pests. The outcome of competition among two or more parasitoid species may be altered if the parasitoids are differentially attacked by the same hyperparasitoids. A reliable assessment of the hyperparasitoid community is needed to understand how top-down trophic interactions influence the effectiveness of introduced parasitoids. We examined the diversity of hyperparasitoids attacking Cotesia glomerata in Colorado (USA), where the congener C. rubecula is absent. We compared this diversity with the hyperparasitoid community of C. glomerata and C. rubecula from Maryland (USA) where both wasps co-occur and use the imported cabbageworm (Pieris rapae) as their main host. Field collected C. glomerata broods were analyzed to examine the relationship between brood sizes and the adult sex ratio and the likelihood of attack by different hyperparasitoid species. A total of nine hyperparasitoid species were found in Colorado, of which four species also occurred in Maryland. While larger C. glomerata broods experienced increased odds of hyperparasitism, C. glomerata developing in larger broods had higher per capita survivorship than those developing in smaller broods. The proportion of adult male C. glomerata in a brood increased with brood size in both non-hyperparasitized and hyperparasitized broods, suggesting that female C. glomerata were not preferentially hyperparasitized. Hyperparasitoids inflicted greater mortality on C. rubecula than on C. glomerata. This differential hyperparasitism may enable the coexistence of C. glomerata with its congener C. rubecula, which usually outcompetes and displaces C. glomerata.