The spread of cassava mosaic disease (CMD) and populations of the whitefly vector (Bemisia tabaci) were recorded in cassava when grown alone and when intercropped with maize and/or cowpea. The trials were planted under conditions of high inoculum pressure in 1995 and 1996 at a site in the lowland rainforest zone of southern Cameroon. In the 1995 experiment, the maize and cowpea intercrops reduced the final incidence of CMD in the cassava cvs. Dschang White and Dschang Violet, but not in the more resistant cv. Improved. In the 1996 experiment with cv. Dschang Violet, the maize and cowpea intercrops grown alone or together decreased adult whitefly populations on cassava by 50% and CMD incidence by 20%. The monomolecular population growth model generally provided the best fit for disease progress. Areas under the disease progress curves (AUDPCs) and incidences expressed as multiple infection units were significantly (P<0.05) less for cassava intercropped with maize and/or cowpea than in cassava alone; times to 50% CMD incidence were significantly (P<0.05) longer in all intercrop systems. In 1995 the basic infection rates (r) were similarly low (0.010 per month) in the moderately resistant cv. Dschang Violet intercropped with maize and in all treatments in the more resistant cv. Improved. By contrast, rates were significantly higherfor cv. Dschang Violet alone or with cowpea and in all treatments for the less resistant cv. Dschang White (0.030???0.060). In 1996, r values in cassava grown alone (0.077) were significantly larger (P<0.05) than in the other cropping systems (0.042???0.052). There were no significant differences in the symptom severity in the different cropping systems. Disease foci were isodametric and more compact in plots containing cowpea than in other cropping systems.