Summary.The Barn Swallow is a non‐breeding winter visitor to West Malaysia (Malaya), abundant in season, by day feeding aerially over a wide range of habitats and by night normally roosting gregariously in trees, reed‐beds or on service wires in towns. Records of ringed birds have demonstrated that those reaching Malaya breed in the Palaearctic region from 108°E eastwards between 37° and 51°N. Recoveries south of the breeding range suggested that migrating birds may follow either a continental route or a more easterly track through the Philippines and Borneo. Counts at roost sites in a reed‐bed and in towns demonstrated a seasonal increase in numbers from late July to a peak in November, followed by a decline of about 20% to a level maintained until mid‐February when departure commenced. Most birds had left by early May, but a few lingered and possibly overlapped with the first returning migrants in June. There was no evidence that any individuals remained in Malaya through the nuptial period. Repeats during winter at three regularly sampled urban roosts indicated that many birds on passage were present until November and again in late March–early April; from December to February the winter population was relatively stable and comparatively sedentary. Although the distances between towns were small in relation to the demonstrated foraging range of Barn Swallows, only 17% of 1,955 repeats of ringed birds represented a shift in roost site. Most shifts were towards the centrally situated and most populous roost of the three; interchanges between the outer pair of towns were few.A complete moult occurred on the wintering grounds, during which young of the year acquired adult plumage. Replacement of the primaries extended virtually throughout the moulting period, at an average rate of 2.4 feathers per month in the proximal part of the tract and 1.3 feathers per month in the distal part. Adults on average moulted slightly earlier than juveniles, but there was a wide scatter in timing between individuals of both age groups. There was no evidence that the initiation of moult was related to the dates of post‐nuptial migration. The date of departure on prenuptial migration, however, was normally delayed until primary moult was complete. Large weight gains in March and April occurred only in swallows which had completed the moult. At this period the mean weight of birds in fresh plumage was about 30% above the lowest winter mean, and was significantly higher than that of contemporary samples of birds in which moult was continuing. In final samples in late April and early May mean weights showed a decline, indicating that late birds departed with reduced deposits of metabolic reserves.The gonads of adults of both sexes among passage and arriving birds in July and August had largely completed post‐nuptial regeneration, and subsequently remained quiescent. Preliminary stages of recrudescence were observed in females from February onwards, and in males from March. Recrudescence was most advance in specimens which had completed the moult, but did not approach breeding condition in any bird before departure.Returning birds tended to be conservative in their choice of winter roost. Among 1,276 records, 82% were recaptured in the town of original ringing. Again shifts towards the centrally situated roost were more numerous than between the peripheral pair. The frequency of returns varied significantly with the month of ringing, being higher for December‐March, lower for July‐November and April‐May. Survival rates, calculated from returns after one and two breeding seasons, indicated an annual mortality of 60–72%, higher among juveniles than adults. Comparison of results of successive years suggested that unfavourable conditions in 1967 resulted in lower survival of juveniles in particular than in 1966. There was no evidence of mortality at the roost sites, and it is argued that heavy losses probably occur durin