Dispersal is a fundamental but poorly understood process in ecology, evolution and conservation. Natal dispersal patterns are a major determinant of population kin structure and thus may play a key role in social evolution. We studied natal dispersal in the cooperatively breeding long-tailed tit Aegithalos caudatus, a non-territorial species that does not exhibit delayed dispersal. We investigated the factors associated with local recruitment and dispersal distance using 11 years of data, and generated dispersal distance distributions (DDDs) before and after correcting for sampling bias. We also examined how dispersal direction varied between and within nests, using a novel randomisation method to correct for bias. Recruitment was male-biased and increased with nestling weight, and there were significant nest and year effects. Neither sex nor weight had a significant effect on dispersal distance, but distance increased with brood size and there was a significant nest effect. The observed DDD was right-skewed but the corrected DDD was almost symmetrical, and this correction more than doubled estimates of mean dispersal distance and fledgling recruitment rate. There was no tendency across nests for birds to disperse in a particular direction, but siblings dispersed in similar directions. These results provide a detailed description of natal dispersal in long-tailed tits with minimal bias. They also demonstrate the importance of studying the different aspects of dispersal in combination, and show that direction is an important component of dispersal that is usually overlooked. The pattern of natal dispersal is consistent with the nature of helping behaviour in long-tailed tits, and may play an important role in the evolution and maintenance of cooperation despite the absence of delayed dispersal.