Musical instrument training has been found to be associated with higher cognitive performance in older age. However, it is not clear whether this association reflects a reduced rate of cognitive decline in older age (differential preservation), and/or the persistence of cognitive advantages associated with childhood musical training (preserved differentiation). It is also unclear whether this association is consistent across different cognitive domains. Our sample included 420 participants from the Lothian Birth Cohort 1936. Between ages 70 and 82, participants had completed the same 13 cognitive tests (every 3 years), measuring the cognitive domains of verbal ability, verbal memory, processing speed, and visuospatial ability. At age 82, participants reported their lifetime musical experiences; 40% had played a musical instrument, mostly in childhood and adolescence. In minimally adjusted models, participants with greater experience playing a musical instrument tended to perform better across each cognitive domain at age 70 and this association persisted at subsequent waves up to age 82. After controlling for additional covariates (childhood cognitive ability, years of education, socioeconomic status, and health variables), only associations with processing speed (β = 0.131, p = .044) and visuospatial ability (β = 0.154, p = .008) remained statistically significant. Participants with different amounts of experience playing a musical instrument showed similar rates of decline across each cognitive domain between ages 70 and 82. These results suggest a preserved differentiation effect: Cognitive advantages (in processing speed and visuospatial ability) associated with experience playing a musical instrument (mostly earlier in life) are preserved during older age.