Summary Bacillus thuringiensis (Bt) crops kill pest larvae but have led to resistance evolution in several target pests. The high dose‐refuge (HDR) strategy aimed at delaying Bt resistance evolution depends on dispersal patterns of target pests. Examination of adult dispersal of the European corn borer Ostrinia nubilalis (ECB), the main target of Bt maize, can help to improve resistance management. Estimated recapture rates over 20 mark–release–recapture sessions in herbaceous field borders, where ECB adults rest during the day and mate at night, were used to examine the influence of sex, release period and site on ECB dispersal. Data from an additional 30 sessions were used to test the influence of night temperature, humidity, dew index and wind speed. Average recaptures within 50 m of release were lower 12 h after night (7·7%) than 12 h after day (34·5%) releases, did not differ between sexes, and decreased during nights with higher temperatures and lower wind speed. Local habitat had a major influence on dispersal. The number of unmarked adults caught initially in a given section of field border was strongly correlated with those subsequently captured in the same section, suggesting that moths flying in from the surroundings consistently settle in the same preferred spots. Moreover, recapture rates of marked adults were positively correlated with the prior density of unmarked adults in the release section. The spatial distribution of recaptured moths around the release point suggests that they moved on a very local scale, while those not recaptured probably left the area by a different, long‐range type of dispersal. Synthesis and applications. A proportion of European corn borer adults typically remained within a few metres of their initial location for at least 12 h. This should favour non‐random mating early in the flight season when nights are cold, population mixture is low and most individuals are unmated. Non‐random mating can accelerate the evolution of resistance, but this effect may be offset by non‐random oviposition. Our findings suggest that the intensity and direction of dispersal could be manipulated by field border management. Our data on the range and prevalence of short‐range dispersal and the factors influencing this process, support the view that resistance evolution is multifactorial. Our results can be used to parameterize multifactorial models from which specific management recommendations can be formulated.