The purpose of this study was to examine associations for work-related stress, reactivity to stress, and subsequent risk of dementia. The sample consisted of members of the population-based Swedish Twin Registry who were participants in the HARMONY study (n=2049). We used case control and cotwin control designs, with information on work-related stress and reactivity to stress collected as part of a questionnaire completed in 1967. Dementia was diagnosed approximately 30 years later using a 2-stage procedure-screening for cognitive impairment followed by full clinical evaluation. We found that measures of work-related stress (job dissatisfaction and high job demands) were not associated with dementia risk. Greater reactivity to stress predicted higher risk of dementia controlling for age, education, sex, occupational status, alcohol use, and smoking status (odds ratio=1.57, 95% confidence interval 1.08-2.31). Cotwin control analyses also showed that dementia probands were more likely to report high reactivity to stress than their nondemented cotwins. We did not find evidence of an interaction between work stress and reactivity in predicting dementia. Overall, indicators of stress due to environment (ie, work) were not associated with dementia, whereas the individual characteristic of reactivity to stress predicted dementia risk. © 2007 Lippincott Williams & Wilkins, Inc.