We studied dispersal movements by sibling pairs of great tits, Parus major, and blue tits, P. caeruleus, in a patchy environment, in order to test whether siblings are more similar in dispersal than expected by chance. Because of possible common environmental effects due to the heterogeneity and finiteness of the study area, we compared the similarity among siblings with the similarity between each sibling and an unrelated bird that fledged in the same patch and year, as close to the siblings' nest as possible. Siblings of both species were not more similar in dispersal distance than they were to control birds. However, great tit siblings dispersed in similar directions compared to control birds, and this result was not affected by the degree of matching between sibling and control birds. As a consequence, siblings ended up breeding at closer distances from one another than control birds. Heritability values calculated from parent-offspring regressions were close to zero, suggesting that there is no additive genetic variance for dispersal distance or dispersal direction. We propose that similarity in dispersal direction originates from association of siblings during dispersal or during activities that influence the choice of direction, such as postfledging family movements. Our results show that non-independence in the choice of dispersal direction by siblings may influence small-scale kin structure in this population with high local recruitment.