Hand represents the main organoleptic characteristics of clothes. Evaluation of this characteristic is based on the base of subjective meanings. The subjective handle is closely connected with the objective measureable surface, mechanical, bulk and thermal properties of textiles. The main aim of this project is specification, the objective characteristics for prediction of subjective hand and creation of predictive type model. The methodology of predictive model building is based on the application of graphically oriented techniques for regression modelling. This methodology is used for evaluation of a predictive model for prediction of the subjective hand of PET/wool fabrics.