Abstract Surveys were carried out in Zambia twice yearly from 1986 to 1990, to assess the impact of a biological control campaign against the cassava mealybug (CM), Phenacoccus manihoti Mat.-Ferr., throughout the infested area. From 1984 onward, the exotic parasitoid Epidinocarsis lopezi (De Santis) (Hym., Encyrtidae) and some exotic coccinellid predators were released on 54 occasions along the spreading front of CM infestation. E. lopezi established in every release site, spread, and covered the entire infested area, but the exotic coccinellids did not establish. Between 1986 and 1990, CM populations declined on average 5.8 times. In a multiple regression analysis involving meteorological, agronomic, plant, and entomological variables, from a total of 4804 cassava fields, nine variables had a significant influence on the CM population density, which in turn was the main factor influencing tip damage scores. The duration of E. lopezi′s presence in an area was the most important factor: in the year the CM was recorded for the first time in a particular district, more than 20% of all cassava fields had an average of 10 or more CM/tip (a few having as many as 1000). This percentage was gradually reduced to 0% in Year 5. The condition of the plant, its age, rainfall, and water retention capacity of the soil were also important. Where CM populations were lower, damage by the cassava green mite, Mononychellus tanajoa (Bondar) (Acari, Tetranychidae), appeared more severe. The frequency of ants increased with the CM population density. In conclusion, biological control of the CM in Zambia was successful.