This paper is about the evaluation of stress response and recovery using biosignals and fuzzy theory. We caused mental stress by means of a coin-stacking task. During the experiment, 4 kinds of biosignals, including frontalis EMG, ECG, peripheral skin temperature and skin conductance level, were acquired. Then, the degree of stress was assessed by synthetically those signals using fuzzy inference. From the fuzzy inference result, the parameters (amount of physiological change / amount of imposed stress) and (time to 25% recovery), which represent response and recovery respectively, were derived. We made a two-dimensional point graph using the response parameter as an abscissa and the recovery parameter as an ordinate for each subject.