BackgroundSleep dysfunction is signature of Alzheimer’s disease and Parkinson’s disease (PD), and can signify incipient disease, disease risk, and worsen symptoms over time. How real‐world sleep dysfunction relates to patient self‐report of sleep and clinical cognitive dysfunction is poorly understood, partly because self ‐report is impaired in patients with cognitive decline. We monitored real‐world sleep with wearable actigraphy devices in patients with PD to test the hypothesis that objective patterns of sleep dysfunction are associated with worse cognitive performance.MethodTwenty‐nine participants with idiopathic PD (age = 67.44 ± 5.79, 20 males) completed the Montreal Cognitive Assessment (MoCA) and Epworth Sleepiness Scale (ESS) to capture clinical cognitive decline and self‐reported daytime sleepiness. Sleep was monitored for 4‐weeks using wrist‐worn ActiGraphs. Sub‐scores and total scores of MoCA were compared with Total Sleep Time (TST), Sleep Efficiency (SE), Wakefulness After Sleep Onset (WASO), and Sleep Fragmentation Index (SFI) measured by actigraphy, as well as ESS total using a Pearson correlation.ResultWorse sleep fragmentation (SFI), the percentage of awakenings and movements during sleep, predicted worse cognitive impairment overall (MoCA score: r = ‐0.38, p < .05) and delayed recall (r = ‐0.44, p < .05). Reduced sleep time (TST) and worse sleep quality (SE, WASO) did not worsen patient cognitive impairment. Patient self‐report of sleepiness (ESS) did not associate with worse cognitive outcomes.ConclusionThis pilot analysis identifies sleep fragmentation as a key risk factor for cognitive dysfunction in PD. Patient self‐report of sleep may not reliably reflect chronic sleep disruption and related cognitive dysfunction. Results underscore that objective measures of real‐world dysfunction can help inform clinical care and intervention for patients at risk for cognitive decline and dementia.