Cardiac 123I-meta-iodobenzyl guanidine (123I-MIBG) uptake, using the mediastinal region as a reference, is a predictor of outcome in heart failure. Constant uptake in a reference region is required when studying the pharmacotherapeutic effects on the integrity of the cardiac adrenergic system. The aim of this study was to verify the assumption of constant uptake in reference regions and their relation to established predictors of clinical outcome. Planar 123I-MIBG images were obtained in 28 patients with heart failure. The left ventricular ejection fraction correlated with the heart-to-mediastinum (H/M) ratio but not with myocardial 123I-MIBG uptake. Plasma noradrenaline correlated with the H/M ratio, the heart-to-lung (H/L) ratio, mediastinal uptake and age, but not with uptake in the myocardium and lung. Age correlated with the H/M ratio, H/L ratio, mediastinal uptake and lung uptake, but not with myocardial uptake. In a regression model, after the inclusion of age, the H/M ratio and mediastinal uptake were not related to noradrenaline. We conclude that changes in the H/M and H/L ratios are predominantly due to confounding changes in mediastinal uptake or lung uptake. In pharmacological intervention studies, myocardial 123I-MIBG uptake should be assessed using single photon emission tomography or positron emission tomography, which do not rely on reference regions in mediastinum or lung.