The issue of differential prediction of health outcomes by sociological models of work stress has received little attention so far. This paper argues, both on theoretical and empirical grounds, that active coping with the experience of chronic work stress is more likely to be associated with physical health consequences of sustained autonomic arousal such as hypertension, whereas passive coping may predispose individuals to withdrawal behavior such as sickness absence from work. Based on data from a cross-sectional study on 189 male middle-aged (40-55, 48.3 +/- 4.6 years) middle managers in a car-producing company in Germany, this hypothesis is tested in the framework of the theoretical model of effort-reward imbalance at work. More specifically, the simultaneous manifestation of high effort and low reward at work, indicative of active coping, is expected to statistically predict the risk of being hypertensive. Conversely, suffering from low occupational rewards in the absence of signs of sustained effort, indicative of passive coping, is expected to predict sickness absence (SA) behavior. Multivariate odds ratios (OR) derived from logistic regression analysis and adjusted for important confounders indicate that three measures of low reward are associated with short-term SA (OR ranging from 3.30 to 9.15), that one measure of low reward is associated with long-term SA (OR: 2.67) and that two measures of low reward are associated with number of SA episodes (OR 4.05 and 6.33), whereas no indicator of high effort at work is significantly associated with SA. On the other hand, the OR of being hypertensive is 5.77 in middle managers who suffer from high effort and low reward simultaneously. In conclusion, a sociological model of work stress which allows for differential prediction of health outcomes according to the important notions of active versus passive coping with work demands finds preliminary empirical support.