Introduction: Even though people with congenital prosopagnosia (CP) never develop a normal ability to "overtly" recognize faces, some individuals show indices of "covert" (orimplicit) face recognition. The aim of this study was to demonstrate covert face recognition in CP when participants could not overtly recognize the faces. Methods: Eleven people with CP completed three tasks assessing their overt face recognition ability, and three tasks assessing their "covert" face recognition: a Forced choice familiarity task, a Forced choice cued task, and a Priming task. Results: Evidence of covert recognition was observed with the Forced choice familiarity task, but not the Priming task. In addition, we propose that the Forced choice cued task does not measure covert processing as such, but instead "provoked-overt" recognition. Conclusions: Our study clearly shows that people with CP demonstrate covert recognition for faces that they cannot overtly recognize, and that behavioural tasks vary in their sensitivity to detect covert recognition in CP.