A cross-sectional study of 3860 health-sector workers across two data collections was conducted to identify the predictive power of different job demands and job resources during the COVID-19 pandemic based on four indicators of distress (COVID-19 traumatic stress, burnout, generalised anxiety, and depression) among health-sector workers. Exploratory and confirmatory factor analyses, measurement invariance checks, and structural equation models were used to evaluate the dimensionality and the effect of the job demands and resources on distress indictors. The identified job demands were workload, confinement, loss, and virus exposure, while the identified job resources were self-efficacy, momentary recuperation, and meaning making. Loss and workload predicted the distress indicators best, while confinement and virus exposure mainly predicted COVID-19 traumatic stress and were less important for the other distress outcomes. Self-efficacy and meaning making negatively predicted distress, while momentary recuperation, controlled for the other demands and resources, was positively related to the distress indicators. Of the typical pandemic-related demands and resources, the experience of loss due to COVID-19 infection was the most important predictor of distress outcomes. Confinement, and especially the awareness of virus exposure, were far less important predictors.