The aim of the present study conducted among 274 Finnish employees was to examine the relationships between job characteristics, recovery experiences and occupational well-being across 1 year. We hypothesized that these relationships would follow normal causation, that is, job characteristics at T1 predict recovery experiences (detachment, relaxation, mastery and control at off-job time) and well-being (fatigue at work and work engagement) at T2, and recovery experiences at T1 predict well-being at T2. The structural equation modelling analyses lent support to the hypothesized normal causation model compared with the reversed causation and reciprocal models. However, only the longitudinal relationships between job characteristics and recovery experiences were significant. More specifically, high job demands at T1 predicted poor detachment, relaxation and control during off-job time at T2; and high job resources at T1 predicted high mastery experiences in off-job time at T2. Thus, our study showed that job demands in particular inhibited recovery experiences in the long term, but this was not reflected in employee well-being across 1 year. Copyright © 2013 John Wiley & Sons, Ltd.