Aims: Despite landmark heart failure (HF) with reduced ejection fraction (HFrEF) trials showing effect of mineralocorticoid receptor antagonists (MRA) on the risk of death and HF hospitalization, it has been suggested that MRAs are underutilized or frequently withdrawn. This study sought to identify temporal trends in the initiation of MRAs and the subsequent risk of withdrawal and adherence of MRAs in HF patients treated with a renin–angiotensin system inhibitor and a beta-blocker in Denmark from 2003–2017. Methods and results: From nationwide registries, we identified patients receiving a diagnosis of HF. Use of MRA was identified by at least one prescription within 6 months after the diagnosis. The absolute risk of withdrawal with treatment was assessed with cumulative incidence, accounting for the competing risk of death. To estimate adherence, we calculated the proportion of days covered. We included 51 512 patients with incident HF. During the study period, 20 779 (40.3%) patients initiated MRA therapy. The incidence of withdrawal of MRA was 49.2% throughout the study period; 48.0% of the HF patients were adherent to the treatment. Among patients withdrawing treatment with MRA, the cumulative incidence of reinitiating was 36.6%. Conclusions: In a nationwide cohort of patients with HF, approximately half of the patients received MRA as third-line therapy within the first 6 months after diagnosis and approximately half of these withdrew MRA within 5 years. These findings warrant an increasing focus on retention to MRA treatment in a real-life setting.