Recent research has suggested that individual differences in the personality measure of hardiness are related to health status. This 3-year longitudinal study with full-time employees (N=229) investigated the association between stress, two separate measures of personality hardiness, and objective (absenteeism due to illness verified from personnel records) and subjective (self-reported hospitalization due to injury/illness and psychological well-being) health measures. Using stepwise multiple regression analysis, stress, but not hardiness, significantly predicted absenteeism when controlling for relevant demographic variables. Stress and the alternative measure of hardiness also significantly predicted psychological well-being. Finally, the alternative measure of hardiness, but not any of the original personality hardiness scales, predicted self-reported hospitalization for injury/illness over the 3-year period. Little evidence for the predictive validity of the Kobasa personality hardiness scales or overall hardiness score existed with either objective or subjective health outcomes in this study. However, the alternate measure of hardiness, but not the original hardiness scales, was significantly associated with well-being in this prospective study. These findings suggest that the current conceptualization and measurement of the original hardiness construct should be re-evaluated. (Author/ABL)