Computer simulation was used to study the evolution of resistance to pesticides under four main classes of conditions common to agricultural arthropod pests and natural enemies. Results show that the influence of biological and operational factors on the rate of resistance evolution depends on the presence or absence of immigration and the functional dominance of resistance. A high-dose strategy for retarding resistance, which requires high immigration by susceptibles and functionally recessive resistance, can disrupt biological control and promote resistance in secondary pests. A low-pesticide-use strategy which can retard resistance in pests and promote resistance in natural enemies is recommended.