Abstract Introduction Sleep spindle activity has been increasingly studied as an underlying mechanism of cognition. In youth, it appears the relationship between spindle activity and cognition depends upon the spindle metric and cognitive domain examined. Prior research has been conducted primarily in highly selective experimental studies of typically developing youth. We aimed to clarify the relationship between spindle activity and lower and higher order cognitive functions in children and adolescents from the general population. Methods We studied 639 children aged 5-12y (median 9y) and 418 adolescents aged 12-23y (median 16y) from a population-based cohort. All subjects underwent a 9-hour, in-lab polysomnography. We calculated sleep spindle density (SSD), the total number of spindles per minute of stage 2 of non-rapid eye movement sleep, and peak spindle frequency (PSF) in the 10-16 Hz range at central, frontal and fronto-occipital derivations. Wechsler intelligence testing assessed verbal and non-verbal intelligence quotients (IQ), processing speed (coding) and working memory (digit span backward [DSB]). Multivariable-adjusted linear regression models with age, sex, race/ethnicity, body mass index, apnea/hypopnea index, and insomnia symptoms as covariates examined the association between SSD and PSF with cognitive outcomes. Results At ages 5-12, central SSD was positively associated with verbal IQ (p=0.04), non-verbal IQ (p=0.03), coding (p=0.01) and DSB (p<0.01); additionally, frontal SSD was positively associated with coding and DSB (both p<0.01) and fronto-occipital SSD with DSB (p<0.01). Also, central (p<0.01) and frontal (p=0.01) PSF was positively associated with DSB. At ages 12-23, fronto-occipital SSD was positively associated with non-verbal IQ (p=0.02), while no other statistically significant associations were observed for SSD or PSF with cognitive outcomes (all p≥0.08). Conclusion Spindle density is a strong correlate of general ability (both verbal and non-verbal IQ) in childhood, and it remains for non-verbal IQ in adolescence. Both increased spindle density and peak frequency are associated with better working memory in childhood, yet not in adolescence. These developmental differences may be due to cortical (e.g., synaptic pruning) and thalamocortical (e.g., increased myelination) maturational changes occurring during adolescence. Support (If Any) National Institutes of Health (R01MH118308, UL1TR000127)