The malaria model previously fitted to 1 year of baseline data from the Garki District in the Sudan savanna of northern Nigeria was tested against data collected in the same area over a period of 3 years, including 1(1/2) years during which the insides of houses in certain villages were sprayed with propoxur. It was also tested against data collected in Kisumu, Kenya, also over a period of 3 years, including 20 months during which the insides of houses in part of the area were sprayed with fenitrothion. The test consisted in using the vectorial capacity, calculated from the entomological observations made in the above places and periods, as input in the Garki model while keeping the other parameters as fitted to the Garki baseline data, and in comparing the prevalence of Plasmodium falciparum parasitaemia as estimated by the model to that actually observed. There was relatively good agreement and the model is considered epidemiologically satisfactory and fit for use in planning malaria control operations.