The walking and flight dispersal of marked overwintered and summer Colorado potato beetles (CPB), Leptinotarsa decemlineata (Say) (Coleoptera: Chrysomelidae), released in field box‐plots was monitored simultaneously in six habitats over a period of 4 days. The emigration out of plots by walking beetles was calculated from the catch in linear pitfall traps completely surrounding each box‐plot and emigration flight was estimated from the number of beetles missing from the plot or captured by the trap. Overwintered beetles dispersed sooner after release than summer beetles. Overall, the mean number of beetles retained by the habitat was significantly higher in the host habitat (potato) than in any non‐host habitat tested (soybean, pasture, bare ground, water, woodland). Unexpectedly, there was no or little difference in overall beetle retention between non‐host habitats except for higher retention in the water habitat. No difference in the ratio of flight over walking could be detected by the study between overwintered and summer CPB except in the water and woodland habitats. Twenty‐four hours after release, the highest ratios were obtained in the water and woodland habitats and the lowest in the bare‐ground habitat, but ratios were similar for all habitats, except water, after 96 h. As a population, under these experimental conditions, 96 h after release, it seems that CPB displayed a slight preference for flight over walking, with walking as a default mode. A fed and starved pre‐release treatment had no effect on dispersal rates or mode of dispersal. Essentially, our results showed that over a 96‐h period, northeastern North American CPB emigrated at similar rates from the various non‐host habitats encountered, except for water, using walking as much as flight. The host habitat retained CPB significantly longer than non‐host habitats but with a mode of dispersal ratio similar to that in non‐host habitats. The impact on dispersal of the various habitats encountered by CPB in the agro‐ecosystem was less important than expected suggesting that the interaction of environmental parameters is likely to have the most significant impact in determining dispersal rates and dispersal modes.