Artificial Face Recognition is one of the popular areas of research in Image processing. It is different from other biometric recognition because faces are complex, multidimensional and almost all human faces have a similar construction. Out of many issues some of the most important issues associated with facial recognition are the type, format and composition (different background, variant illumination and different facial expression) of the face images used for the recognition. Different approaches use the specific databases which consist of single type, format and composition of image. Thus, these approaches lack the robustness. So, in this paper an comparative study is made for four different image databases with the help of basic Eigenface PCA face recognition technique.