Changes in sedentary behaviours and physical activity according to retirement status need to be better defined. Retirement is a critical life period that may influence a number of health behaviours. We assessed past-year sedentary behaviours (television, computer and reading time during leisure, occupational and domestic sitting time, in h/week) and physical activity (leisure, occupational and domestic, in h/week) over 6 years (2000-2001 and 2007) using the Modifiable Activity Questionnaire in 2,841 participants (mean age: 57.3±5.0 y) of the SU.VI.MAX (Supplementation with Antioxidants and Minerals) cohort. Analyses were performed according to retirement status. Subjects retired in 2001 and 2007 (40%) were those who spent most time in sedentary behaviour and in physical activity during and outside leisure (p<0.001). Leisure-time sedentary behaviours increased in all subjects during follow-up (p<0.001), but subjects who retired between 2001 and 2007 (31%) were those who reported the greatest changes (+8.4±0.42 h/week for a combined indicator of leisure-time sedentary behaviour). They also had the greatest increase in time spent in leisure-time physical activity (+2.5±0.2 h/week). In subjects not retired 2001 and 2007 (29%), changes in time spent watching television were found positively associated with an increase in occupational physical activity (p = 0.04) and negatively associated with changes in leisure-time physical activity (p = 0.02). No consistent association between changes in sedentary behaviours and changes in physical activity was observed in subjects retired in 2001 and 2007. Public health interventions should target retiring age populations not only to encourage physical activity but also to limit sedentary behaviours.