Job stress is commonly defined in terms of role demands originating in the work environment (Cooper and Marshall, 1977; Carayon, 1992). An increasing body of research suggests that a major reaction produced by job stressors is burnout (see Lee and Ashforth (1993) and Cordes and Dougherty (1993) for a review of the early research). Since Freudenberger (1974) coined the term ‘burnout’ to describe a state of chronic emotional fatigue, this phenomenon has been the focus of much research interest. The most widely accepted definition of burnout conceptualises the phenomenon as ‘a syndrome of emotional exhaustion, depersonalisation and reduced personal accomplishment that can occur among people who do ‘people work’ of some kind’ (Maslach, Jackson and Leiter, 1996). This reflects the fact that early research work was conducted in the ‘caring’ professions. It is now recognised that burnout can and does occur in a wide range of occupations and is not unique to the human service sector. For example, Dolan (1995) studied burnout among senior executives, Bacharach, Bamberger and Conley (1991) undertook a comparative study of burnout in samples of engineers and nurses and Cordes, Dougherty and Blum (1997) investigated burnout among human resource managers and professionals in corporate settings. In response to the recognition that burnout is not unique to human service workers, Maslach and her colleagues (1996) have developed a new version of their burnout inventory (the MBI) designed to measure burnout outside the human services sector (the MBI-GS). This instrument has been found to possess good factorial validity and satisfactory internal consistency reliability across a range of different occupational groups and countries (Schutte et al., 2000).