We study the impact of lifelong job tasks and occupational stressors on health in older age. Using SHARE data, we link occupational stressors to 4-digit ISCO occupations
and rely on a rich set of control variables. Physical exertion, measured by metabolic rate consumption, is the most impactful occupational stressor. The effects are larger for
women. Our results remain robust to Oster bounds and the Lewbel IV approach. We track how occupational strain contributes to long-term health outcomes by studying
disease progression over the life-cycle, highlighting gender heterogeneity. In the second part of the paper, we develop a LASSO-based approach to identify harmful tasks and construct a Job Strain Intensity measure accounting for unobserved stressors. This framework can be used as a tool whenever specific tasks play a crucial role in policymaking. For instance, it can help the design of targeted retirement schemes or improve safety measures for high-risk tasks.