The present study allows for a more flexible Job Demand–Control model by introducing years of work experience as a contextual factor (i.e. moderator). Building on the processes of adaptation and diminishing returns to learning-by-doing, the authors assumed that the relationships between job demands, job control and strain versus learning decrease with the number of years an individual has been working (i.e. years of work experience). Hypotheses were tested using data from Flemish workers during their first years on the labour market (N = 3158). The authors accounted for unobserved individual heterogeneity by means of panel data techniques. Results revealed that the impact of control on both strain and learning decreased with years of work experience (i.e. main effect). The authors also found a significant three-way interaction effect between demands, control and years of work experience for both strain and learning, in line with their expectations.