Although a link between depression and poor adherence to antihypertensive medication (AHM) has been found, it is not known whether depression actually leads to poorer adherence or whether poor adherence only is a marker of depression. In this study we aimed to determine the time order between hypertension, depression and changes in adherence to AHM. The analyses were based on data gathered from a longitudinal cohort of Finnish employees (The Finnish Public Sector Study). A total of 852 chronically hypertensive men and women at baseline with a recorded onset of depression during the 9-year observation window and 2359 hypertensive control participants matched for age, sex, socio-economic status, time of study entry, employer and geographic area were included in the study. Individuals with any sign of depression during 4 years before the beginning of the study were excluded. To describe long-term trajectories (4 years before and 4 years after the recorded depression) of AHM adherence in relation to the onset of depression, annual data on reimbursed AHM prescriptions were gathered from the national Drug Prescription Register. Annual nonadherence rates (i.e. number of 'days-not-treated') were based on filled prescriptions. Among male cases, the rate of 'days-not-treated' was 1.52 times higher (95% confidence interval 1.08-2.14) in the years after the onset of depression compared to preonset levels. In women and in male controls, no change in adherence to AHM was observed between these time periods. In hypertensive men, the onset of recorded depression increases the risk of nonadherence to AHM.