Stemborers are major pests of maize and sorghum in tropical Africa. Various management approaches to stemborer control are being sought with emphasis on those that are environmentally sound, sustainable and economically feasible. Studies reported in this thesis were conducted to explore the potential of employing habitat management techniques to enhance the effectiveness of stemborer parasitoids. Emphasis was placed on the chemical and behavioural ecology of stemborer parasitoids in a diversified system that consisted of the stemborer host plants, maize and sorghum, stemborers and a non-host plant, molasses grass ( Melinis minutiflora ).Olfactometric studies on the stimuli involved in long-range foraging behaviour of the parasitoids Cotesia sesamiae and Dentichasmias busseolae revealed that host plant volatiles play a major role in guiding searching parasitoids to habitats that harbour their hosts. Volatiles from herbivore-injured plants were the most attractive. Sorghum was more attractive to the parasitoids than maize. Molasses grass volatiles were attractive to C. sesamiae but repellent to D. busseolae . Combining host plants and molasses grass did not have an additive effect of increasing the response of the parasitoids. Local growth conditions influenced the volatile blend produced by molasses grass grown in two different locations in Kenya, Thika and Mbita. This was manifested by their differential attractiveness to C. sesamiae . Dentichasmias busseolae did not discriminate between host species at a distance based on the volatiles released by infested host plants. Headspace analyses of volatiles emitted by uninfested and infested host plants and molasses grass grown at Thika and Mbita, revealed qualitative differences in their compositions. Infested host plants released a richer volatile blend than the uninfested host plants. Most of the compounds identified in the herbivore-injured plants were previously reported to play a role in parasitoid recruitment. Molasses grass from Thika had a number of its identified compounds that were similar to some in the blend of infested host plants. These compounds might have played a role in attracting C. sesamiae . Molasses grass from Mbita had very few compounds common to the other plants.Semi-field studies on the close-range searching behaviour of C. sesamiae showed that the parasitoid was not arrested by molasses grass although the grass had previously been demonstrated to be attractive to this parasitoid. The presence of molasses grass in a patch with host plants did not interfere with the foraging behaviour of D. busseolae . Both parasitoids recognized host plants at close range and were arrested on infested host plants where they spent most of their time engaged in host searching activities. Field studies showed that stemborer densities were lower in the intercrop than in the monocrop, while diversifying the habitat with molasses grass had no effect on stemborer parasitism. Larval and pupal parasitism was 2.1 and 11.0% in the monocrop compared to 2.0 and 9.8% in the intercrop.The main conclusion of this thesis is that intercropping maize or sorghum with molasses grass does not enhance the foraging behaviour of stemborer parasitoids. Molasses grass seems to be of more importance to the herbivore than the parasitoids. Hence, to clearly understand why stemborer densities were lower in the intercrop than the monocrop, future studies should focus on the interrelations between molasses grass, stemborers and host plants.