Several aspects of the migratory behaviour of birds, notably timing, direction and duration, have been shown, mainly from breeding experiments, to be under genetic influence. Nevertheless, aspects of migratory behaviour can be improved by learning, either from personal experience or from other individuals. This review is concerned with social influence on different aspects of bird migration. Communal migration can provide not only the usual anti-predation and food-finding benefits, but also energy savings (resulting from particular flock structures), greater synchronization of individual journeys and improved route finding. Individuals in a group can pool their knowledge of directions and routes to mutual benefit, and if a group contains different age-groups, naïve youngsters can learn from more experienced individuals. These advantages may be one reason why many birds migrate in flocks or in more diffuse aggregations, even some species that normally live solitarily. Some species, notably swans, geese and cranes, migrate as family units within flocks, so young could learn migratory routes, stopover sites and wintering areas from their parents. And providing the two age-groups migrated together, the young of other species could learn the same details from older, more experienced individuals, whether related or not. In the daytime, some birds can be seen to migrate on their own, while others travel in groups, flocks or other looser assemblages. Night migration has been studied by radar, ceilometers and other artificial lights and by 'moon-watching', as well as by sound recording to detect flight calls. While some species travel at night in dense flocks, as in the daytime, others (including some solitary species) seem to travel in loose aggregations which may form once the birds are airborne. Calling maintains contact between individuals and may thereby provide information on optimal directions, flight altitudes and other aspects. Species vary in the extent to which they call at night, but such vocalizations increase at times of change, such as take-offs and landings, altitude changes or as birds enter mist. Many birds show specific behaviours before migration which serve to synchronize departures of those individuals which have reached an appropriate condition. Individual homing pigeons, released at a distance, can normally find their way back to their home loft, but different pigeons may develop different routes. When pairs of pigeons, which have developed different routes, are released together, they may either take an intermediate route or one pigeon may join the other regarded as 'leader'. The larger the difference in experience between two pigeons, the greater the likelihood that the more experienced bird will emerge as leader. Migration in flocks may improve the accuracy of orientation if the direction of the flock is an average of the headings of the individual flock members (the 'many wrongs' principle). This notion has been supported by experiments on homing pigeons, and also by observations of wild birds, in which single individuals or small groups showed more variation in their directions than did larger groups. In general, directional scatter decreased as group size increased. In such systems, individual birds benefit from social contact with others, but all individuals have equal influence on the resulting direction. In practice, some birds (leaders) are likely to have more influence on directions than others (experienced more than inexperienced), as shown in experiments with pigeons and tracking of wild goose families on migration. Large scale field experiments have also shown that birds can be influenced in their movements by others of their species, and that even birds from resident populations can sometimes be induced to migrate when released into areas where migrants prevail. When juveniles are released in the absence of experienced adults, they take a wider range of directions than normal, some of which take them off an optimal route. By training geese, swans or cranes to fly behind ultralight aircraft, new migration routes have been created, re-establishing migratory populations in areas from which they were eliminated in the past. Such experiments further illustrate the role of social influence on bird migratory patterns. Suggestions are made for further research aimed to fill gaps in understanding and to take forward our understanding of social influence on migratory behaviour. [ABSTRACT FROM AUTHOR]