One of the most important characteristics of comfort is hand. It is well known that hand plays important role as the first characteristics entering to contact with consumer. Evaluation is carried out by consumer on the basis of his feeling evoked by textile with contact of his preceptors (fingers and palms). With development of new types of technologies and textile products the objective characterization of hand becomes more important. This together with transferring to computer-oriented methods leads to finding indirect but objective techniques for subjective hand evaluation based on the special regression types (multivariate calibration). The heavy cotton fabrics used as military or protective textiles have often-unsatisfactory hand and other comfort parameters. The main challenge is therefore to improve their hand by the proper construction or finishing. The main aim of this contribution is utilization of well-known KES (Kawabata Evaluation System) for creation of regression type prediction model suitable for estimation of the heavy cotton fabrics total hand value (THV). The set of properties correlated strongly with total hand value are selected. The proposed model is used for prediction of softeners influence on improving of THV and fabrics selected mechanical properties.