Progress curves of cassava mosaic virus disease (CMD) and populations of the whitefly vector (Bemisia tabaci) were assessed using four cassava varieties grown alone and as a random mixture in two experiments established under epidemic conditions at a site near Kampala in southern Uganda. There were significant differences in final CMD incidence and in the areas under the disease progress curves between varieties when grown alone and as a mixture in both experiments. Variety Ebwanateraka had the highest incidence and SS4 the lowest, even though it supported the largest populations of adult whiteflies. The overall incidence of CMD in the mixture was similar to that in pure stands of the partially resistant Nase 2 and greater than in the resistant Migyera and SS4. Compared to pure stands, incidence of CMD in each component of the mixture was reduced significantly only in Ebwanateraka, whereas vector populations were less only in SS4 and Nase 2. On several observation dates the actual incidence of CMD and populations of adult whiteflies in the mixture were significantly less than expected values estimated from the results for the four varieties when each was grown alone. A highly significant positive relationship was established for each variety between peak populations of adult whitefly and leaf area index at the time. The implications of the findings and the scope for future research on the use of varietal mixtures for the management of CMD are discussed.