We study the impact on earned income of living in neighbourhoods characterised by a special type of residential segregation, called social frontiers. Possible labour market outcomes of living in neighbourhoods marked by social frontiers have not been previously studied. Social frontiers represent abrupt transitions between adjacent neighbourhoods and arise when there are significant and substantial spatial divisions or "cliff-edges" in the proportion of population groups in the contiguous neighbourhoods. This type of segregation, where highly different demographic groups are located close to each other, could cause reduced inter-group contact, which may impact labour market attainment. We use rich individual-level administrative data from Oslo (Norway) and advanced statistical methods for systematically identifying social frontiers along immigrant-native lines. Our causal design applies propensity score matching (PSM) to control for self-selection into neighbourhoods, which might otherwise bias the results. We find that residing on the immigrant dense side of a social frontier has statistically significant negative effect on the earnings of men, but overall, the effects are small.