Discrimination in hiring against certain demographic subgroups can be an explanatory factor of the high levels of unemployment in those groups. The present study proposes an evaluation of the cross effects of gender and ethnic origin on the access to employment of fictitious candidates with identical productive characteristics. This study is based on our own experimental data obtained by the paired testing method. We made up 8 fictitious job applications from young graduates with a master's degree in computing, and sent them simultaneously to the same 303 job offers between the beginning of February and the beginning of April 2009. The only differences between these fictitious candidates were their sex and their ethnic origin, as signalled ...