The extent to which the impact of computer skills depends on how computers are used is investigated using British data from an establishment survey, cohort studies and the European E-Living survey. We examine the importance of activity and frequency of use in these various data sources. We find that the impact on earnings depends on which cohort of workers is examined and that there are differences over time. The regression results show that the use of computers for internet access and for email is positively significant across all of our datasets, although there are differences in the size of the effects between men and women.