We study the relationship between age and literacy skills using data from the IALS, ALL and PIAAC surveys. In cross-sectional data there is a negative partial relationship between literacy skills and age that is statistically significant indicating that literacy declines with age, especially after age 45. However, this relationship could reflect some combination of age and birth cohort effects. In order to isolate age effects, we use the three international surveys to create synthetic cohorts. Our analysis shows that in most participating OECD countries the negative slope of the literacy-age profile in cross-sectional data arises from offsetting ageing and cohort effects. In these countries more recent birth cohorts have lower levels of literacy and individuals from a given birth cohort lose literacy skills after they leave school at a rate greater than indicated by cross-sectional estimates. Finland, Italy and the Netherlands are exceptions to this pattern; in these countries more recent cohorts have higher literacy levels and the cross-sectional estimates overstate the rate at which literacy declines with age. Our birth cohort results suggest that there is not a general tendency for literacy skills to decline from one generation to the next, but that the majority of the countries examined are doing a poorer job of developing literacy skills in successive generations.