Alterations in sleep patterns are common among older adults; further, short and long sleep durations have been linked with impaired cognitive performance in older individuals. Yet most research examining these relationships has been cross-sectional, limited to high-income nations, and has failed to consider how changes in sleep duration may impact cognitive decline. The present longitudinal study uses nationally-representative data to test whether changes in sleep length among "healthy" baseline sleepers are associated with reduced cognitive function in older Mexican adults (>50 years old) at follow-up. Data were drawn from the first and second waves of the World Health Organization's Study on global AGEing and adult health. Self-report data captured sleep duration over two nights, and five cognitive tests (immediate and delayed verbal recall, forward and backward digit span, and verbal fluency) were used to measure various cognitive domains and create a composite z-score of cognitive performance. Linear regressions were performed to assess associations between sleep length changes and cognitive decline, controlling for relevant lifestyle and health factors. Increased sleep durations at follow-up among individuals who reported intermediate sleep durations (6-9 h/night) at baseline were significantly associated with greater rates of decline in overall cognitive function. Longer sleepers also trended toward greater rates of decline for attention/working memory and executive function. This study suggests that long sleep durations are a risk factor for certain types of impaired cognition among older adults living in a middle-income country. These findings are clinically important given the growing rates of dementia and aging populations globally.