A case-control designed was used to investigate associations and interactions between muscle activity measured by surface electromyography (EMG) in the upper trapezius muscle and subjectively reported risk factors in workers with and without shoulder and neck pain. EMG data were collected both in the workplace (indicating vocational muscle activity) and in a laboratory setting (indicating non-specific muscle activity). Women in manual (15 pairs) and office (24 pairs) work were included. The pairs were matched on age, gender and on current and historical work load, such as working hours, type and length of employment. Previous reports of this study have indicated that shoulder and neck myalgia was associated with increased muscle activity for the manual workers, and with psychological and psychosocial factors for the office workers. These risk factors were in the present report used as the basis for studying associations and interactions between muscle activity (1), psychological and psychosocial factors (2), and shoulder and neck pain (3). Subjectively reported or perceived general tension, a stress symptom presumed related to psychosocial and psychological factors, was previously found to be the strongest and only variable separating cases and controls in both work groups. In this paper, no relationship was found between perceived general tension and EMG variables for the office workers. For the manual workers a strong interaction was found; perceived general tension correlated positively with EMG variables for the controls, and negatively with EMG variables for the cases. It is hypothesised that the feeling of general tension represents a physiological activation response that may or may not include muscle fibre activation. This implies that pain provoked by psychosocial stress factors may not be mediated through increased muscle activity.