Adrenomedullin (ADM) correlates with adverse cardiovascular outcomes in patients with acute myocardial infarction (AMI) and in patients with heart failure. Measurement of human mature ADM (mADM) has been difficult, and recent studies have used its surrogate - the mid-regional pro-ADM (MRproADM). Our objective was to determine whether mADM measured by a novel sandwich immunoassay, using the anti-C-terminal and an anti-mid-regional monoclonal antibody, was prognostic of 30-day, 90-day, 1-year, and 2-year major adverse cardiovascular events (MACEs) in 1111 consecutive patients who have suffered an AMI. We also compared it with the effect of MRproADM in the same patient population. A total of 311 (27.0%) patients experienced the primary endpoint at 2 years follow-up. The median (inter-quartile range) of mADM was significantly higher in patients who experienced a 2-year MACE [60.90 (44.00-86.97)] pg/ml, compared to event-free survivors [49.59 (36.20-68.15)] pg/ml (P < 0.001). mADM, taken as 1 SD of the continuous variable, was predictive of MACEs in multivariate analysis, with hazard ratios [95% confidence intervals (CIs)] at 90 days [1.28 (1.01-1.62)], 1 year [1.31 (1.08-1.59)], and 2 years [1.42 (1.07-1.64)]. It was also independently predictive of death at 1-year [1.52 (1.12-2.05)] and 2-year [1.42 (1.07-1.89)] follow-up. mADM was a better predictor of these outcomes than MRproADM, apart from death at 90 days, and combined death and heart failure hospitalization at 1 and 2 years, respectively. Human mADM can be reliably measured and predicts MACE events at medium-term follow-up, and confirms the paradigm of risk stratification using MRproADM - a surrogate for the active hormone. The relationship between mADM and MACE appears to be a continuum.