One of the major challenges in managing the Colorado potato beetle, Leptinotarsa decemlineata (Say) is its remarkable ability to develop resistance to virtually every insecticide that has ever been used against it. Resistance is particularly common throughout northeastern USA as far north as Maine. The first instances of resistance to imidacloprid have already been reported from several locations in New York, Delaware and southern Maine. Rotating insecticides with different modes of action may delay insecticide resistance, but successful implementation of this technique depends on a good understanding of resistance and cross‐resistance patterns in populations of target pests. LC₅₀ values were measured for imidacloprid and thiamethoxam in Colorado potato beetle populations from a variety of locations in the USA and Canada using diet incorporation bioassays. The field performance of imidacloprid, thiamethoxam and clothianidin against imidacloprid‐resistant beetles on a commercial potato farm in southern Maine was also evaluated. Correlation between LC₅₀ values for imidacloprid and thiamethoxam was highly significant, even when populations previously exposed to thiamethoxam were excluded from the analysis. There was no statistically detectable difference in the LC₅₀ values between populations exposed to both insecticides and to imidacloprid alone. Applications of neonicotinoid insecticides at planting delayed build‐up of imidacloprid‐resistant beetle populations on field plots by 1–2 weeks but failed to provide adequate crop protection. Consistently with bioassay results, there was also substantial cross‐resistance among the three tested neonicotinoid insecticides. Results of the present study support the recommendation to avoid rotating imidacloprid with thiamethoxam as a part of a resistance management plan. Copyright © 2006 Society of Chemical Industry