Abstract Introduction Alterations in sleep and circadian rhythms are common in persons with Alzheimer’s disease (AD) dementia, but the nature of such changes in the early phases of AD remains unclear. This study compared sleep and circadian rest/activity rhythms (RARs), measured by standard and novel actigraphic indices, between participants with normal cognition or mild cognitive impairment (MCI), and examined cross-sectional associations between these measures and cognition. Methods Actigraphy data were collected in 179 individuals (mean age=72.6 years, gender=64.8% female) with normal cognition (n=153) or MCI (n=26) from the Biomarkers for Older Controls at Risk of Dementia (BIOCARD) study. Standard sleep parameters (i.e., total sleep time [TST], sleep efficiency [SE], wake after sleep onset [WASO], average wake bout length [WBL]), and standard non-parametric RAR metrics (i.e., relative amplitude [RA], intradaily variability [IV], interdaily stability [IS]) were generated. Functional principal component (fPC) methods were used to generate three novel RAR indices (fPC1, fPC2, fPC3) representing 69% of the total variance. Cognitive test scores were used to generate composite measures reflecting the domains of episodic memory and executive function using factor analysis. Regression models were used to compare sleep and circadian RAR parameters between the diagnostic groups and to evaluate their associations with cognitive performance. Results After adjustment for age, sex, education, and APOE-4 genotype, compared to normal controls, MCI subjects had significantly lower SE, lower RA, and lower scores on the novel RAR measure fPC3, which reflects a later rhythm phase, lower amplitude, and lower activity both at night and early in the day. In analyses combining data from participants with MCI and controls, several standard RAR parameters (e.g., higher RA and IS) and higher fPC3 scores were associated with both better episodic memory and executive function. Additionally, several standard measures (e.g., lower WASO and IV) and lower fPC1 scores (reflecting higher rhythm amplitude and greater activity throughout daytime hours) were linked with better executive function. Conclusion MCI participants have sleep and circadian alterations, which are significantly associated with cognitive performance. A novel RAR measure, fPC3, showed differences in rhythm patterns that extended from the night into the daytime. Support (if any) Funding-support NIA (U19-AG033655, T32-AG027668, R01-AG050507) and AASMF (#223-BS-19).