This paper describes the demands, supports-constraints framework originally articulated by Payne (1979) for predicting psychological distress, and reports the results of ail empirical study involving 2452 white-collar, public sector employees in Australia. The study uses hierarchical regression to test for the interaction effects of the demands and constraints variables, having applied suitable controls for instance by including trait anxiety and removing curvilinear effects. The results reveal little evidence of an interactive effect but moderately strong main effects and modest support for the value of controlling for curvilinear effects. Trait anxiety is shown to have a strong effect on psychological distress and considerably reduces the size of the relationship between demands, supports-constraints and psychological distress when statistically controlled fur. However, it still accounts fur 18.7% of the variance when entered last in the regression and it is recommended that its effects be explored in all studies of stress that rely oil self-report data.