The development of modern robotics has triggered increasing interest in developing artificial tactile sensory systems. However, the tactile mechanism in the design process has been highly limited to experimental tests that are expensive and time-consuming. This work is concerned with the development of using virtual tests for tactile sensory response prediction and design, which includes numerical simulation settings, database creation, response regression and interpolation, and sensor sensitivity designs. Experimental verifications and numerical demonstrations are performed based on the NUS NeuTouch sensors. The potential of using virtual tests to design new tactile sensors to improve response linearity is illustrated. Display omitted •Tactile response prediction and design using virtual tests is proposed with demonstration on NUS NeuTouch sensors.•The framework consists of numerical simulations, database creation, response regression and interpolation.•The potential of using virtual tests to design new tactile sensors for improving response linearity is illustrated.