The overall aim was to investigate the effect of psychosocial loads on trapezius muscle activity during computer keying work and during short and long breaks. In 12 female subjects, surface electromyography (EMG) was recorded bilaterally from the upper trapezius muscle during a standardized one hand keying task—interspaced with short (30 s) and long (4 min) breaks—in sessions with and without a combination of cognitive and emotional stressors. Adding psychosocial loads to the same physical work did not increase the activity of the trapezius muscle on either the keying or the control side, both of which remained at median and static EMG activity levels of around 5% and 2.5% of the maximal voluntary electrical activity (EMGmax), respectively. The difference between the keying and the control side was significant; and further the control side activity was significantly increased above resting level. During both short and long breaks, exposure to psychosocial loads also did not increase the activity of the trapezius muscle either on the side of the keying or the control hand. Of note is that during long breaks the muscle activity of the keying side as well as that of the control side remained at the same level as during the short breaks, which was increased above resting level. This was to be seen from the static and the median EMG activity levels as well as gap times, the overall mean values being: 0.4%EMGmax, 1.1%EMGmax, and 50% in gap time, respectively. In conclusion: psychosocial loads are not solely responsible for increased non-postural muscle activity; and increasing the duration of breaks does not per se cause muscle relaxation. [ABSTRACT FROM AUTHOR]