- Objective: Abnormalities in sodium homeostasis are common in hospitalized patients. Hyponatremia upon admission is a poor prognostic marker in acute myocardial infarction (AMI) patients. However, little is known about the association between changes in sodium levels and in-hospital mortality. We delineated changes in sodium levels and studied the association of such changes with in-hospital mortality of AMI patients. Methods: Retrospective analysis of AMI patients hospitalized for > 6 days. Sodium levels throughout the 6-day post-admission were divided into five equally sized groups (quintiles = Q) and thereafter categorized as follows: Q1 (< 135 mEq/L), Q2-Q4 (135-140 mEq/L, reference group), and Q5 (≥141 mEq/L). Primary outcome: in-hospital mortality. Results: A total of 8306 patients (10,416 admissions) were included (mean age 67.8 ± 14.0 years, 33.4% women, 45.5% STEMI). In-hospital mortality was 6.6%. Q1 and Q5 upon admission were both related to higher risk for in-hospital mortality, compared with the reference group (OR 1.47 and OR 1.33, respectively, p < 0.001 each). Q1 was more frequent in non-survivors throughout the entire study period, while the prevalence of Q5 levels was similar in survivors and non-survivors upon admission carrying increasing mortality risk thereafter: for Q1 consistent OR 1.50, while for Q5 it, increased from OR 1.32 upon admission to OR 1.90 on the sixth day, p < 0.001. Conclusions: Low and high sodium levels are associated with increased risk for in-hospital mortality in patients with AMI. The risk is unchanged for hyponatremia, while it consistently increases for increased sodium levels.