An experiment was conducted to investigate the effect of key stiffness on the development of fatigue, keyboard reaction forces, and muscle electromyography (EMG) responses. Six subjects typed continuously for 2 hours on each of two keyboards (0.28 N or 0.83 N resistance keys, presented in random order). Keyboard reaction force and root mean square finger flexor and extensor EMG were recorded for 2 minutes at 250 Hz for every 10 minutes subjects typed. After typing for 2 hours subjects were given a 2-hour rest break and then typed on the remaining keyboard for an additional 2 hours Fifty-four percent more peak force, 34% more peak finger flexor EMG, and 2% more peak finger extensor EMG were exerted while using the 0.83 N keyboard. Peak and 90th percentile values showed similar trends and were well correlated for force and finger flexor and extensor EMG. Subjects typed much harder than necessary (4.1 to 7.0 times harder on the 0.28 N keyboard and 2.2 to 3.5 times harder on the 0.83 N keyboard) to activate the keys. Fatigue was observed on the 0.83 N keyboard during 2 hours of continuous typing, but the trends were mild. It appears that the ratio of typing force to flexor EMG may not be a sensitive enough indicator of fatigue for low-force high repetition work.