Despite research focused on the control of E. saccharina Walker (Lepidoptera: Pyralidae), it remains the most destructive pest of sugarcane in South Africa and therefore a shift in the approach to the control of this insect was required. Habitat management techniques were employed through trials to understand the mechanisms used by insect pests in their host plant oviposition behaviour. Indigenous and beneficial non-crop plants, which could be used to attract insect pests away from sugarcane were identified and tested. An African grass, Melinis minutiflora, shown to be repellent to lepidopteran stemboring pests, was used in field trials in sugarcane. Eldana saccharina was shown to be fairly indiscrimate in choosing host plant species. Oviposition trials showed that females made no choice for host plants oviposition based on the volatiles released by those plants. Females showed no preference for males over test plants. But did consistently move and make a plant choice more often than male moths. Moths were not attracted by the volatile stimuli of a host plant and the availability of cryptic sites might be a factor that influenced ovipositing females to choose a host plant. Field trials tested the repellent action of Melinis minutiflora against E. saccharina and were shown to be more effective over a big field with space for M. minutiflora to establish thick undergrowth alongside a field plot. The other field sites showed no significant effect from M. minutiflora intercropped into treatment plots, or a slight negative effect. Later sugarcane planting times in relation to the grass planting time was a possible reason for the positive result in only two field sites allowing the grass to grow and establish before the sugarcane competed with the grass for sunlight. A cost benefit analysis of planting a hectare of sugarcane with M. minutiflora showed an economic benefit linked to reduction in E. saccharina infestation. Field sites with low population pressure from E. saccharina would not yield the economic benefit of planting this grass. There was no significant loss in the height, density or sucrose yield (ERC% cane) between control and treatment plots in the field plots due to the presence of M. minutiflora. Comparison of weed biomass between treatment and control plots showed a significant reduction in the treatment plot where M. minutiflora out-competed the weeds already present. This grass was advantageous in the sugarcane field as it sometimes reduced moth infestation but did not significantly compete with sugarcane and showed weed suppressing potential.