Purpose <p>Exiting paid work can be associated with mental health changes. Antidepressant medication is an indicator for mental health, but its trajectories have mostly been studied in younger populations. However, the trajectories could be different among older adults, who are a rapidly growing population group. Therefore, our aim was to explore these trajectories among people who continued working until 65–70 years and whether sociodemographic factors and prior sickness absence (SA) and disability pension (DP) due to mental diagnosis (mental SADP) were associated with such trajectories.</p> Methods <p>All residents of Sweden aged 65–70 years old in 2010, who were in paid work 2005–2011 and exited paid work 2012–2016 (<i>N</i> = 32,849, 39% women), were included. We used register microdata for five years before and after work exit and included sociodemographic factors at the exit year and mental SADP in the years prior to exit. We used group-based trajectory modelling to examine trajectories in antidepressant medication (ATC-code N06A) and multinomial logistic regression to study factors associated with each trajectory.</p> Results <p>Altogether, 90% had no antidepressant medication. Among those with medication, 94% had constant and 6% increasing use. Mental SADP before work exit was the strongest determinant of belonging to both identified trajectories. Older age at exit was associated with a lower likelihood of belonging to the increasing use trajectory.</p> Conclusion <p>Antidepressant medication patterns remained largely constant after work exit. Prior mental SADP was the most consistent determinant of antidepressant medication among people exiting work at an older age.</p>