Abstract BACKGROUND Pathophysiological studies suggest that A-type natriuretic peptides (ANPs) might provide valuable information beyond B-type natriuretic peptides (BNPs) about cardiac dysfunction in patients with coronary heart disease (CHD). We aimed to assess the predictive value of midregional pro–A-type natriuretic peptide (MR-proANP) for recurrent cardiovascular disease (CVD) events in stable CHD patients for whom information on N-terminal proBNP (NT-proBNP) was already available. METHODS Plasma concentrations of MR-proANP and NT-proBNP were measured at baseline in a cohort of 1048 patients aged 30–70 years with CHD who were participating in an in-hospital rehabilitation program. Main outcome measures were cardiovascular mortality, nonfatal myocardial infarction, and nonfatal stroke. RESULTS During a median follow-up of 8.1 years, 150 patients (incidence 21.1 per 1000 patient-years) experienced a secondary CVD event. MR-proANP was associated with a hazard ratio (HR) of 1.89 (95% CI, 1.01–3.57) when the top quartile was compared to the bottom quartile in the fully adjusted model (P for trend = 0.011). For NT-proBNP the respective HR was 2.22 (95% CI, 1.19–4.14) with a P for trend = 0.001. Finally, MR-proANP improved various model performance measures, including c-statistics and reclassification metrics, but without being superior to NT-proBNP. CONCLUSIONS Although we found an independent association of MR-proANP as well as NT-proBNP when used as single markers with recurrent CVD events after adjustment for established risk factors, the results of a simultaneous assessment of both markers indicated that MR-proANP fails to provide additional prognostic information to NT-proBNP in the population studied.