Purpose – The purpose of this paper is to investigate the link between childhood neighbourhood ethnic composition and short- and long-run economic outcomes of second-generation immigrants and natives in Sweden. Design/methodology/approach – The author uses Swedish longitudinal register data and apply regression analysis methods to investigate the correlation between three ethnic neighbourhood variables(share of immigrants, share of immigrants with the same ethnic background and share of immigrants with other descent) in childhood with short- and long-run economic outcomes (earnings, unemployment, reliance on social assistance and educational attainment). Findings – The results show that second-generation immigrants raised in immigrant-dense neighbourhoods have a lower probability to continue to higher education, whereas, their earnings, unemployment and social assistance tendencies are unaffected. On the contrary, natives’ earnings and educational attainment are negatively correlated with, and the probability of social assistance and unemployment are positively associated with a high immigrant concentration. Moreover, the social assistance and unemployment of non-Nordic second-generation immigrants appears to be negatively correlated with the neighbourhood share of co-ethnics and positively correlated with the neighbourhood proportion of other ethnic groups. Overall, the author finds that the results are very similar in the short and long run. Originality/value – This paper expands the literature on children and ethnic segregation and in contrast to earlier research in this context, it focuses on second-generation immigrants and their performance in comparison to natives. This study contributes to this research area by investigating a large variety of outcomes, looking at both immigrant, own ethnic group and other ethnic group concentration and including both short- and long-run correlations. [ABSTRACT FROM AUTHOR]