While there is ample evidence supporting genetic control of migratory behaviour in short-lived passerines, long-lived social species have been assumed to rely solely on cultural inheritance of migratory routes. Evidence from experimental studies supporting this idea is scarce. We tested whether the moult migration in taiga bean geese Anser fabalis has an inherited component or whether the birds need oblique social learning (where knowledge on migration is transferred from any experienced individual to any naïve individual conspecific) to carry out this journey. In many waterfowl species, non-breeders and failed breeders migrate to remote places for wing moult while successful breeders stay at the breeding grounds and moult with their chicks. We translocated one-year-old taiga bean geese before their first moult migration to sites outside of the breeding range to examine whether they display innate moult migration behaviour without experienced conspecifics or not. The birds were equipped with GPS-transmitters and released in randomly assigned groups of two. Wild control one-year-old birds were released immediately after capture with other non-breeding geese, while a procedural control group consisting of older birds was held in captivity until released at the same time with the translocated one-year-old birds but in the place where they were captured. Most translocated birds found conspecifics and either joined locally moulting breeders or followed experienced birds to moulting sites in Russia. Two of the translocated birds did not find other bean geese and settled to moult together in SW Finland. The wild control birds moult-migrated as expected, while only one of the procedural control birds moult-migrated to Russia and the remaining three stayed with locally moulting breeders in Finland. Our results support the idea that moult migration in geese is culturally inherited, highlighting the importance of the non-relative, experienced adult individuals have in maintaining population-specific behaviours.