This work presents the main advantages of the use of longitudinal data for the sociological research on the assimilation process. Two fundamental issues will be developed: the advantages panel data offer in controlling for unobserved individual heterogeneities on one hand, and, on the other hand, their capacity to make some advances in the estimation of causal effects. Two empirical examples related to each of these advantages are provided by analysing inequalities between immigrants and natives in the French labour. These examples put the stress on some methodological issues linked to the use of such data.