||In this paper we analyse the data obtained in the automatic regime by the Optical Electronic Device (OED, Vorotkov et al. 2009; Bolshakov et al. 2010) for autumn nocturnal passage of passerines on the Courish Spit on the Baltic Sea and estimate: (1) numbers aloft under differ- ent types of wind (following wind, opposing wind and calm conditions); (2) flight trajectories in the 5Â° cone of white light. We found that under natural nocturnal illumination conditions, the vertical cone of white light impacts the detectable numbers aloft and disturbs straight flight trajectories. The OED data obtained throughout the night suggest, after correction for ground speed and the mean flight altitude, the actual number of birds in the light cone peaks at calm conditions, is halved under following winds which are optimal for passage and is 21 times lower under unfavourable headwinds. It is assumed that high numbers in the light cone under calm conditions is an artefact of bird attraction by light and their concentration around the searchlights. The OED data obtained for midnight Â±1 hour, flying migrants respond to the vertical light cone under all types of wind conditions by altering their straight flight trajec- tories. However, this response is most apparent in still air conditions. The proportion of birds that change their flight track reaches 43%. We assume that under such conditions some birds are not only attracted to the illuminated zone at low altitudes, but, besides slowing down their ground speed, change their trajectories to the degree of flying in circles. To determine combi- nations of factors and to test for their possible impact on the probability of response to light, we used a binary logistic regression. The presence of birds with straight vs. curved tracks was used as the dependent variable. Final logistic models obtained for midnight Â±1 hour for calm conditions and headwinds, suggest that occurrence probability of songbirds with curvilinear flight tracks is higher for small birds, when no or just a small part of Moon disk is visible and under high air humidity. Under headwinds the probability of occurrence of birds flying curvi- linear tracks is also higher under overcast. For following winds, the probability of occurrence birds flying curvilinearly was higher when many small birds were aloft, when air humidity was high and when wind was not strictly following. Unlike other wind situations, this model did not include the size of visible part of the Moon disk as a significant factor. The increase of occurrence of curvilinear flight tracks through the light beam when winds were not exactly following was probably caused by the problems with compensating the lateral component of tailwinds under high velocities, especially by small birds.