So-called master faces are faces for which the probability of achieving a false match when being compared against an unknown set of face images is significantly higher than that of a random impostor. In this work, we investigate whether average faces exhibit said property of master faces. In order to generate average faces for different demographic groups, face morphing is applied to images of public face databases. Subsequently, average faces are compared against different databases in a biometric identification scenario considering various demographic groups. Obtained results indicate that for some demographic groups, e.g. Asian females, average faces tend to yield moderately higher chances for a false match, while generally this is not the case.