We performed a correspondence testing in order to assess the potential discrimination at job access level against young people of ethnic origin from the underprivileged suburbs of the Paris area (Ilede-France). We measure simultaneously the effects of place of residence (privileged or underprivileged city), of nationality (French or Moroccan), and of sound of surname and of forename (French or Arab), on the chances of obtaining a job interview when answering a job ad. We base our assessment on a controlled experiment conducted on the profession of waiter. We constructed 16 jobseeker profiles and sent 938 resumes in reply to 118 job vacancies advertised at the end of 2006. We obtain two results. First, there is evidence of a significant effect against the candidates with an Arab origin; second, there is evidence of residential discrimination against the candidates that are either the most qualified or of French origin. Overall, discrimination would tend to level down the employment opportunities of the candidates in underprivileged suburbs by putting at a disadvantage the candidates that are usually the most favored.