This paper presents a methodology of material property determination by a human subject?s active finger tapping. Object digitization by direct finger touch is a prerequisite for good quality intuitive haptic input interface. Previous attempts were limited to scanning the surface geometry, and were not capable of describing the material characteristics of the object being touched. In this research, a compound mechanical model was developed to characterize the non-linear, viscoelastic behavior of the fingertip and the test object. Using a new active touch digitizer, subject?s tapping input and response were analyzed for material recognition. Parameters to describe material properties were obtained by successive, optimized model fit. In an experiment, our methodology was successful in recognizing the elastic properties of five common materials, including a biomaterial.