This 10-yr field study examined prolonged dormancy (dormancy of more than 1 yr) in Colorado potato beetles, Leptinotarsa decemlineata (Say), from Upstate New York. The research focused on whether the life cycle of the Colorado potato beetle could reduce the effectiveness of crop rotation as a pest management tactic. One set of experiments quantified variation in the occurrence and duration of prolonged dormancy in a natural population. A second set of experiments tested the effect of artificial selection for increased incidence of prolonged dormancy. Of the ≈12,600 unselected beetles in the field cages overall survival (= % emergence) averaged 56.5 ± 18.4% (mean ± SD, range = 23.5–84.3%, n = 19 cages). Most (97.7%) of the emerging beetles surfaced after one winter; the remainder (2.3%) emerged after more than one winter in dormancy. The range of variation in the incidence of prolonged dormancy among the experimental cages during the 10-yr period was 0–7.2%. The incidence of emergence from prolonged dormancy was higher among beetles from the first summer-generation [mean ± SD = 3.0 ± 2.0% (10 cages)] than from the second [1.0 ± 1.5 (7 cages)], and significantly more females than males emerged from prolonged dormancy (female:male = 97:64). Most (≈70%) of the beetles that underwent prolonged dormancy emerged after two winters in diapause; the majority of the remainder (≈29%) emerged after 3–7 yr in diapause; and, one beetle (≈1%) emerged after 9 yr. Artificial selection for prolonged dormancy over three generations did not result in a significant increase in the incidence of delayed emergence, nor did it result in incidences of prolonged dormancy greater than those in the controls (unselected field-reared beetles and lines selected for 1-yr dormancy). From our results, we conclude that the Colorado potato beetle has considerable variation in both the incidence and duration of prolonged dormancy, and that this variation has strong environmental and genetic determinants. Viewed in the context of the beetle’s life history, these findings lead us to suggest that the evolution of resistance to crop rotation would require strong and sustained selection over large areas.