Intra-personal space modeling proposed by Moghaddam et. al. has been successfully applied in face recognition. In their work the regular principal subspaces are derived from the intra-personal space using a principal component analysis and embedded in a probabilistic formulation. In this paper, we derive the principal subspace from the intra-personal kernel space by developing a probabilistic analysis of kernel principal components for face recognition. We test this new algorithm on a subset of the FERET database with illumination and facial expression variations. The recognition performance demonstrates its advantage over other traditional subspace approaches.