Economic integration of immigrants is a process which, like linguistic integration, takes time. Both processes influence each other. In this thesis, we examine the effect of linguistic competence in the official languages of immigrants in Quebec on their economic performance in a longitudinal perspective. We distinguish immigrants according to their mother tongue which, as ail ethnic attribute, might impact on their positioning on the labour market. In the particular case of immigrants and according to the human capital theory, knowledge of the official languages used in Montreal should affect positively their socio-economic situation given that language skills enhance both the communicative competence directly and the transfer of human capital from the country of origin to the host country indirectly. These two elements are known to determine economic performance. Having a different mother tongue than French or English identifies an immigrant as a member of a minority ethnic group which, according to discrimination theory, could lead to stigmatisation and a lesser economic performance. It is possible that this membership explains differential access to information about the job market, the work conditions and the way to make a place on the labour market. The study of these diverse aspects of language will allow us to evaluate its impact oil the economic performance of immigrants. Analyses are carried out on a cohort of immigrants who arrived in Montreal in 1989 and whose main settlement acts were followed during ten years. We selected four moments during the 10 years period in order to examine the economic performance on a short, medium and long term perspective. Socio-economic status conferred by employment as well as the hourly wages are measures of the economic performance. We have repeated and detailed measures of language competence in French and in English as well as the language of education if classes were taken during the observation period. Using multiple linear regressions, we study the explanatory force of different linguistic aspects. Our results suggest that language characteristics of immigrants play a much less significant role in the determination of the hourly wage than anticipated by the theories. The measure of socio-economic status is more sensitive to the impact of language than wages. As the socio-economic status partly determines wages, the influence of linguistic skills on salary is indirect. Knowledge of the official languages gives access to jobs with a higher prestige during the first two years of stay in Quebec only. English has a more positive influence on economic performance than French. Mother tongue identifies immigrant groups with difficult access to higher status jobs and salaries during the 10 years period of observation. We identify particularly immigrants from Haiti, Vietnam and South America as disadvantaged ethnic groups during the whole settlement period. These immigrants work in low prestige jobs and, in these jobs, earn significantly lower salaries compared to anglophone immigrants. Thus they are doubly disadvantaged. Economic integration is not achieved for all immigrants after a 10 year period. Language competence does not seem to be at the origin of social cleavages among immigrants when one takes into account the evolution of their linguistic abilities in French or in English as well as the origin of the language skills. Mother tongue, on the other hand, explains observed status and wage differences among certain immigrants groups during the whole observation period. Measurement of the impact of language competence on economic performance of immigrants is only precise in longitudinal surveys, which follow a group of immigrants, ideally on the same labour market. 'Key words.' Immigration; wages; social status; mother tongues; language competence; language use; Montreal; longitudinal survey.