Background Depression is one of the most prevalent psychiatric disorders, and in older persons is associated with high levels of comorbidity and under-treatment. Dysfunction of the hypothalamic-pituitary-adrenal (HPA) stress axis is consistently observed in the older population as well as depressed patients, with the angiotensin converting enzyme (ACE) a key regulator of the stress response. Epigenetic regulation of ACE may play an important role in HPA axis (dys)regulation. Objective To investigate ACE promoter methylation as a biomarker of late-life depression, and its association with genetic variation and cortisol secretion. Method The longitudinal general population ESPRIT study is aimed at investigating psychiatric disorders in older persons (n = 1863, average age = 73). Depression was assessed using the Mini International Neuropsychiatric Interview according to DSM-IV criteria and the Centre for Epidemiologic Studies Depression Scale (CES-D). Genotype information for seven polymorphisms across the ACE gene was also available. Blood and saliva samples collected at baseline and used to extract DNA and measure cortisol, respectively. Sequenom MassARRAY was used to measure promoter DNA methylation of the ACE gene (n = 552). Results There was no evidence of an association between ACE promoter methylation and depression. However, there was evidence that ACE genetic variants influenced methylation, and modified the association between depression and methylation (Δ at various sites; ‐2.05% to 1.74%; p = 0.019 to 0.039). Multivariate analyses were adjusted for participants’ lifestyle, health and medical history. Independent of depression status, ACE methylation was inversely correlated with cortisol levels (r = −0.336, p = 0.042). Conclusion This study provides evidence that associations between ACE methylation and depression are genotype-dependent, suggesting that the development of reliable depression biomarkers may need to consider methylation levels in combination with underlying genetic variation. ACE methylation may also be a suitable biomarker of cortisol and/or HPA axis activity.