In multimodal biometric technique strength of unimodal system biometrics are combined to increase recognition accuracy of the system. In this paper palm print and finger print biometrics are used to design robust recognition system. A feature of palm print are extracted with gray level co-occurrence based Harlick features and feature of finger print are extracted with minutiae based techniques. The proposed system is tested on publically available IIT Delhi Touch less Palm print database and FVC 2002 database for finger print. In this paper matching score level fusion technique with weighted sum rule based fusion is applied to fuse their individual score of palm print, finger print traits. The recognition accuracy is improved and it is found good as compared with recognition accuracy of individual traits. The multimodal system is evaluated on the basis of performance parameter, Accuracy 99.93 %, and Equal Error Rate (EER) is 0.0006.