BACKGROUND: Hyponatremia is associated with an increased risk of mortality in patients with heart failure and in acute ST-segment elevation myocardial infarction (STEMI). The aim was to assess the impact of hyponatremia on admission on long-term mortality of patients with first ever STEMI or non-STEMI (NSTEMI). DESIGN: This was a longitudinal observation study METHODS: The study population consisted of 3558 patients, aged 25-74 years, with an incident acute myocardial infarction (AMI) in the years 2000-2008 who survived for at least 28 days. All consecutive patients were registered through the Cooperative Health Research in the Region of Augsburg (KORA) Myocardial Infarction Registry. Serum sodium levels were obtained on admission. The association with long-term-mortality was examined using Cox regression models. RESULTS: Hyponatraemia, defined as a sodium level less than 136 mmol/l, was present in 658 (18.5%) patients on admission. During a median follow-up period of six years (interquartile range (IQR) 4.0-8.2 years), 526 patients (14.8%) died. Hyponatraemia was significantly associated with long-term mortality by an 83% higher risk in the age- and sex-adjusted analysis. After further adjustment for reduced left ventricular ejection fraction (LVEF), glomerular filtration rate, haemoglobin, hypertension, hyperlipidaemia, any recanalization therapy, diabetes, medication with diuretics and angiotensin-converting enzyme (ACE) inhibitor/angiotensin-receptor blocker before admission and other parameters hyponatraemia remained a strong predictor for higher long-term mortality (hazard ratio 1.61; 95% confidence interval 1.32-1.97). CONCLUSIONS: Patients with incident AMI and hyponatraemia on admission showed a significantly higher risk of long-term mortality than patients without. This strong predictive value was independent of a number of prognostic factors, including diabetes, glomerular filtration rate or reduced LVEF.