We investigated the development of literacy (reading competence) and numeracy (mathematical competence) across three to six years of adulthood. Data came from two large-scale multi-wave surveys: the PIAAC-longitudinal Study (PIAAC-L, N = 2779) and the National Educational Panel Study (NEPS; N = 3140). We examined mean-level change (ΔT₁, T₂) and rank-order consistencies (rT1, T2) in the total population and in sociodemographic subgroups defined by age, gender, and education. To account for measurement error, we employed plausible values (PV) methodology. Results revealed that literacy and numeracy are highly but not perfectly rank-order stable (.61 ≤ r ≤ .85 in the total population). Mean-level change was negligible for both skills and studies, but there were considerable individual differences in change. Apart from moderate mean-level gains in literacy in young adults (18–29 years) and losses among the highly educated (in NEPS), there were few subgroup differences in mean-level change or rank-order consistency, and most differences did not replicate across studies. Our findings suggest that adults' literacy and numeracy are malleable throughout adulthood and can change even over a three- to six-year period. Individual difference in change dominate the picture and warrant further investigation. Our findings can serve as a benchmark against which to compare future longitudinal findings.