Increasing numbers of studies show that high mental demands of paid work are significantly associated with a slower cognitive decline and reduced risk of dementia/Alzheimer’s disease. Among them, the job complexity framework, particularly complexity with data and people, was very successful in explaining how mentally stimulating jobs contribute to cognitive aging. We propose that novel information processing may be a specific component of mental demand that compensates age-related cognitive decline, which complements the job-complexity framework. Occupation-level job characteristics from the Occupational Information Network are used to operationalize novel information processing at work and linked to individuals’ detailed occupational history in the Health and Retirement Study (N=4,252). We test replicability with MIDUS. Using growth curve modeling, we examine the association of novel information processing at work with cognitive function over fourteen years, we also assess whether this relationship varies by education and across different stages of the life span.