This paper proposes a personal authentication method by fusing face and palmprint recognition for a mobile device. The proposed method is divided into two stages. The first stage recognizes face images by using Verilook 4.0 face recognition system. In the second stage, matching scores obtained from the face and the palmprint recognition developed for this work are fused at score level for the cases yielding low recognition scores in the face recognition. This paper evaluates three different fusion methods: sum rule, weighted sum rule and product rule. Experimental results show that the recognition rates using all three fusion approaches are higher than 98.41%, and the weighted sum rule gives the best performance.