Abstract Among several natural enemies introduced to combat the cassava mealybug, Phenacoccus manihoti (Homoptera: Pseudococcidae), the neotropical parasitoid Apoanagyrus (Epidinocarsis) lopezi (Hymenoptera: Encyrtidae) was the most successful. It established in 26 African countries, causing a satisfactory reduction in the population density of P. manihoti in most farmers' fields. Four conclusions concerning the possible application of the research results to other biological control projects are discussed. (1) Foreign exploration was intensive and should be maintained at this level in other projects, if necessary at the cost of other activities. (2) In the controversy about the amount of research results needed before first releases are made, an understanding of the proper role of quarantine is essential. Whereas quarantine (preferably outside the continent) guarantees nonnoxiousness of natural enemies, only research in the experimental release sites can determine whether a given natural enemy will be efficient. It was confirmed that the released exotic insects did not affect the diversity of the indigenous fauna. Modalities used in this project for the execution of releases, i.e., always on request by and in collaboration with national programs, are recommended for adoption in future projects. (3) Laboratory and field studies established the scientific basis for quantifying the impact of the pest insect and its control by A. lopezi. This was expressed as reduction in pest population levels and yield loss and gain in revenue. Behavior of adult females in searching and choosing hosts was identified as a better predictor of efficiency of a species in the field than life table studies under controlled temperatures. (4) It is concluded that biological control is the basis for integrated pest management. Other interventions, such as cultural methods or the use of resistant varieties, need to be in harmony with biological control because the impact of natural enemies cannot usually be manipulated by the farmer. To achieve sustainability, the aim is to optimize tritrophic interactions among the plant, the phytophagous pest organisms, and their natural enemies, rather than to maximize the effect of a single intervention.