The aim of this note is to propose a modeling approach for assessing the effectiveness of pest biocontrol by natural enemies in diversified agricultural landscapes including several pesticide-based management strategies. Our approach combines a stochastic landscape model with a spatially-explicit model of population dynamics. It enables us to analyze the effect of the landscape composition (proportion of semi-natural habitat, non-treated crops, slightly treated crops and conventionally treated crops) on the effectiveness of pest biocontrol. Effectiveness is measured through environmental and agronomical descriptors, measuring respectively the impact of the pesticides on the environment and the average agronomic productivity of the whole landscape taking into account losses caused by pests. The effectiveness of the pesticide, the intensity of the treatment and the pest intrinsic growth rate are found to be the main drivers of landscape productivity. The loss in productivity due to a reduced use of pesticide can be partly compensated by biocontrol. However, the model suggests that it is not possible to maintain a constant level of productivity while reducing the use of pesticides, even with highly efficient natural enemies. Fragmentation of the semi-natural habitats and increased crop rotation tend to slightly enhance the effectiveness of biocontrol but have a marginal effect compared to the predation rate by natural enemies.