Fatigue is a state of reduced physical activity with a distinctive feeling of discomfort and desire for rest caused by excessive physical and mental activity or illness. Until now, fatigue has been detected by listening to subjective fatigue levels or by measuring reactive oxygen species in the blood, but there is a need for a method that can immediately and easily measure fatigue. In this study, a fatigue task was created on a tablet device and administered continuously for 120 min to induce a temporary neurological state. We recorded the study participants’ voices before and after the fatigue task and examined whether their neural fatigue could be detected using an analysis method based on chaos theory. The analysis showed that cerebral exponent macro (CEM) values, which indicate brain arousal, decreased significantly after the task, except in cases in which concentration on the task seemed to be insufficient．