The question of whether biological control is most likely achieved by deploying single or multiple species of biological control agents is much debated. While utilizing several natural enemies may enhance control, there is also the potential for disruptive inter-specific interactions. Such interactions may be studied in the laboratory by focusing on the details of the interactions themselves and attempting to infer population level consequences from their sum, or by focusing more directly on the overall effects on natural enemy populations: we term these approaches ‘reductionist’ and ‘holistic.’ Here we conduct a holistic laboratory study on interactions between three species of parasitoid wasps that are parasitoids of the coffee berry borer, Hypothenemus hampei (Ferrari) (Coleoptera: Scolytidae): Cephalonomia stephanoderis Betrem, C. hyalinipennis Ashmead and Prorops nasuta Waterston (all Hymenoptera: Bethylidae). We find evidence for both intra- and inter-specific resource competition. Interactions between C. stephanoderis and P. nasuta, both indigenous to Africa, appear to be approximately symmetrical, while C. hyalinipennis, naturally found in the coffee plantations of Chiapas, Mexico, may exert a disruptive influence. C. hyalinipennis also has a low population growth rate. We now consider it to be a detrimental invader of the Mexican coffee agro-ecosystem that should not be encouraged by augmentative release or introduced into other regions. Overall, the most successful species, in terms of both emergence and female production, was P. nasuta. We compare these results with those from prior reductionist and holistic studies, and with observations on patterns of establishment of these bethylid species in the field. Given that it is increasingly clear that disruptive inter-specific interactions are generally common when multiple species are deployed in biological control, screening of potential agents should consider such interactions alongside the more ‘traditional’ focus on host specificity.