Physiological measurements of stress are believed to help with the 'objectivation' of the rather subjective phenomenon called 'stress'. Because there are only moderate interpersonal correlations between the subjective experience of work-stress and physiological stress measurements, the latter are useful mainly for comparing groups with high stress and groups with low stress and not as a diagnostic tool at individual level. Physiological measurements have added value because they quantify the strain that stress exerts on the body, which ultimately determines what effect stress has on health. In addition, physiological measurements on persons with stressrelated symptoms such as burn-out can help us understand the nature of the dysregulation of the stress-regulatory system which occurs in the persons with these types of symptoms. The comparison of high-stress groups and low-stress groups on the basis of various cortisol parameters has produced rather inconsistent results. This seems to be due not only to the unreliability of the measurements but also to the fact that cortisol is apparently a rather inaccurate reflection of the stress-regulatory system in the brain. Research into the changes that occur over time in the cortisol-regulatory system resulting from chronic work-stress and into the health-relevance of various cortisol parameters is essential if progress is to be achieved in this area.