The strong adverse selection that immigrants face in hosting labour markets may induce them to adopt some behaviours or signals to modify employers’ beliefs. Relevant mechanisms for reaching this purpose are personal reputation; exploiting ethnic networks deeply-rooted in the hosting country; and high educational levels used as an indirect signal of productivity. On this last point, the immigrant status needs a stronger signal compared to that necessary for a local worker, and this may lead the immigrant to accept job qualifications which are lower than those achievable through the embodied educational level. This could explain the over education problem that characterizes many countries, Italy included. The aim of the paper is to investigate whether the above mentioned mechanisms are adopted by immigrants in Italy, a crucial country for EU immigration flows, and if they are useful in increasing immigrants’ likelihood of employment. The empirical analysis has been conducted using the dataset from a national Labour Force Survey which provides information on 6,860 documented immigrants. We estimate a logit model for immigrants’ likelihood of being employed, focusing on the above mentioned mechanisms: reputation, ethnic networks and educational level. Moreover we concentrate on the interaction effects of the mechanisms and investigate whether one of them wins on the others. Results show that each of the three mechanisms is statistically and economically significant and exerts positive influence: all factors contribute to increasing the immigrant’s probability of being employed. Anyway, a high level of education increases the probability of being employed more than the belonging to ethnic networks deeply-rooted in Italy. The specific embodied capital of workers matter relatively more. This is relevant for labour public policies in this specific realm since the human capital lever is a possible direct target of various public policies and private human capital investments.