Increasing evidence suggests that circadian and sleep parameters influence cognitive function with aging. To evaluate observational studies of sleep duration and cognition in older adults. A systematic review of OVID Medline and PsycINFO through September 2015, and review of bibliographies from studies identified. English-language articles reporting observational studies of sleep duration and cognitive function in older populations. Data extraction by 2 authors using predefined categories of desired information. Thirty-two studies met our inclusion criteria, with nearly two-thirds published in the past 4 years. One-third of studies indicated that extreme sleep durations were associated with worse cognition in older adults. More studies favored an association with long vs. short sleep durations (35 vs. 26% of studies, respectively). Four studies found that greater changes in sleep duration over time were related to lower cognition. Study design and analytic methods were very heterogeneous across studies; therefore, meta-analysis was not undertaken. We reviewed English-language manuscripts only, with a qualitative summary of studies identified. Observational studies of sleep duration and cognitive function in older adults have produced mixed results, with more studies suggesting that long (rather than short) sleep durations are related to worse cognition. Studies more consistently indicate that greater changes in sleep duration are associated with poor cognition. Future studies should be prospectively designed, with objective sleep assessment and longer follow-up periods; intervention studies are also needed to identify strategies for promoting cognitive health with aging.