Past theory and research indicate that conditions of work can have lasting effects on job incumbents. Karasek and Theorell (1990), for example, proposed that workers' feelings of mastery increase with levels of job demands and job control, and that these effects are mediated by the process of active learning. To test these propositions, 657 school teachers completed scales assessing job demands, control, active learning and mastery on two occasions, eight months apart. As hypothesized, job control predicted change in mastery, an effect that was mediated by active learning. Job demands had a weaker effect on change in mastery. The demands-mastery relationship was moderated by job control, such that under conditions of high but not low control, increasing job demands were associated with gains in mastery. The findings partially support Karasek and Theorell's predictions regarding the main, interactive and mediated effects of job conditions on employee mastery.