This article is the first to examine how 1st-generation migrants affect the employment of workers born in the host country according to their origin, distinguishing between natives and 2nd-generation migrants. To do so, we take advantage of access to a unique linked employer-employee dataset for the Belgian economy enabling us to test these relationships at a quite precise level of the labour market, i.e. the firm level. Fixed effect estimates, including a large number of covariates, suggest complementarity between the employment of 1st-generation migrants and workers born in Belgium (both natives and 2nd-generation migrants, respectively). Several sensitivity tests, considering different levels of aggregation, workers' levels of education, migrants' region of origin, workers' occupations, and sectors corroborate this conclusion.