Evaluation of plant diversity as an insect-pest control strategy. It is widely accepted that more diverse agroecosystems are less affected by insect-pests. There is an abundant body of experimental evidence that supports this belief, but there are also experimental results that contradict it. In order to evaluate plant diversity as a pest control strategy, we reviewed experimental studies published in CAB international between 1998 and 2003. We concentrated on the studies of cases where greater diversification did not lead to more insect-pest regulation. Thirty-three percent of the studies do not support the general assumption of an effective pest control in diversified agroecosystems, which could be explained by the following factors: 1) differing insect life cycles; 2) indirect effects of non–host vegetation on resource quality; 3) surrounding landscape, spatial and temporal scales, and insect population dynamics. Although plant diversity in agroecosystems is generally a convenient strategy for pest control, it is also necessary to evaluate specific designs for each case, due to the simultaneous effect of different variables.