That birds migrate in the reverse direction of the expected is a phenomenon of regular occurrence which has been observed at many sites. Here we use three different methods; tracking radar, radiotelemetry and ringing, to characterize the flights of these reverse migrants and investigate possible causes of reverse migration of nocturnally migrating passerines during autumn migration at Falsterbo peninsula, Sweden. Using these different methods we investigated both internal factors, such as age and fuel load, and external factors such as weather variables, competition and predation risk. Birds flying in the reverse direction were more likely to be lean and to be juveniles. Reverse migration was also more common with overcast skies and winds with north and east components. We did not find any effect of temperature, visibility, number of migrating sparrowhawks, or the total number of ringed birds at the site on the day of departure. We found that reverse migration is characterized by slower flight speeds (airspeed) at high altitudes and that it takes place later in the night than forward migration.During migration it is common to observe birds migrating in the reverse direction of what would be expected for the season. However, not much is known about these reverse movements, and previous studies have given contradictory results as to what causes them (Alerstam 1978, Richardson 1982, Åkesson et al. 1996b. Falsterbo peninsula is a migratory hotspot in southern Sweden, where millions of migrating birds pass each year. Coastlines are in many cases poor stopover habitats, especially at hotspots where many migrants pass, leading to high competition and high predation risk (Alerstam 1978, Woodworth et al. 2014, see also Stach et al. 2015. Poor stopover habitats may increase the amount of reverse migration, and birds have been shown to perform reverse movements to a higher degree at coastal stopover sites such as Falsterbo compared to inland stopover sites (Åkesson 1999).We investigated the factors that distinguish forward from reverse migrating birds during autumn migration at Falsterbo peninsula to gain insights into what causes reverse migration and what characterizes reverse flights compared to forward migratory movements. Both internal factors relating to the state of the bird, and external factors such as weather variables could potentially affect if a bird continues forward or makes reverse movements (Sandberg and Moore 1996). To get an as broad understanding as possible, we combined information about the passage of migrants at the Falsterbo peninsula from tracking radar, radiotelemetry and ringing recoveries. By combining these three methods, we investigated possible causes, both internal and external factors. We also characterize the flight behavior during reverse migration. The possible causes and characteristics of reverse migration that we aim to investigate in this study are summarized in Table 1. Below we describe the predictions to be tested concerning these internal and external factors as well as characteri...