Negative cross-resistance (NCR) toxins that hitherto have not been thought to have practical uses may indeed be useful in the management of resistance alleles. Practical applications of NCR for pest management have been limited (i) by the scarcity of high toxicity NCR toxins among pesticides, (ii) by the lack of systematic methodologies to discover and develop such toxins, as well as (iii) by the lack of deployment tactics that would make NCR attractive. Here we present the concept that NCR toxins can improve the effectiveness of refuges in delaying the evolution of resistance by herbivorous insect pests to transgenic host plants containing insecticidal toxins. In our concept, NCR toxins are deployed in the refuge, and thus are physically separated from the transgenic plants containing the primary plant-protectant gene (PPPG) encoding an insecticidal toxin. Our models show: (i) that use of NCR toxins in the refuge dramatically delays the increase in the frequency of resistance alleles in the insect population; and (ii) that NCR toxins that are only moderately effective in killing insects resistant to the PPPG can greatly improve the durability of transgenic insecticidal toxins. Moderately toxic NCR toxins are more effective in minimizing resistance development in the field when they are deployed in the refuge than when they are pyramided with the PPPG. We explore the potential strengths and weaknesses of deploying NCR toxins in refuges.