Although the relationship between EMG levels and tension headache remains unclear, elevated frontalis EMG levels in tension headache sufferers are usually assumed to be part of the debilitating experience of headache. Elevated tension levels are seldom considered in an activation-type framework. Consequently, headache patients are encouraged to relax in almost all situations. Two experiments were performed to examine whether tension headache sufferers showed any benefits from elevated tension levels during periods demanding a sustained level of cognitive efficiency. Headache subjects were compared with non-headache subjects during discrimination and reaction time task performance across three levels of frontalis tension. Both groups' performance improved under increasing tension conditions relative to the low tension condition. In both experiments improvement was noted as a reduction in the most extreme scores, reflecting possible “mental blocks“. In Experiment 2 median reactions times were also affected. In Experiment I the variation in reaction time scores across the tension levels was greater in the headache group than the non-headache group. Even when the optimal level of tension varied across task difficulty levels, low frontalis tension was never the optimal condition. It was concluded that tension headache prone individuals being treated by relaxation or biofeedback techniques need to be informed about the increased tendency towards slower responding under conditions of frontalis relaxation.