Conservation biological control avoids the need for mass releases of costly agents and the risks associated with introducing exotic agents by promoting existing natural enemies. This is done by alleviating insecticide-induced mortality and by manipulating the habitat to provide resources such as nectar, but there is a dearth of information on the relative and interactive effects of these two approaches. Here we used a large-scale factorial experiment with plots comprised of entire fields to test the effects of, and interactions between, withholding insecticides and planting borders of sesame (Sesamum indicum) on natural enemies and pests over 2 years. We used yellow sticky traps, sweeping netting and sentinel bait plants to monitor natural enemies and pests in the canopy and basal zones of the rice crop. Numbers of rice planthopper egg parasitoids and lepidopterous egg parasitoids in the rice canopy, as well as planthopper parasitism rates, were significantly greater in plots that were unsprayed and bordered by sesame, and scarcest in sprayed crops without sesame. Spraying of sesame-bordered crops gave parasitoid numbers similar to sprayed crops without sesame. Spiders in the canopy were significantly reduced in numbers by spraying, but there was no main effect of sesame borders. This study demonstrates that withholding insecticides and sowing nectar plant borders each have measurable as well as additive benefits on in-crop densities of ecosystem service providers responsible for predating and parasitising pests but the identity of the natural enemy determines the impact of these management practices.