Background: Chronic kidney disease (CKD) patients are prone to both hypo- and hypernatremia. Little has been published on the epidemiology of hypo- and hypernatremia in ambulatory patients with non-dialysis CKD. Methods: Data collected in two contemporaneous CKD cohort studies, the Renal Research Institute (RRI)-CKD study (n = 834) and the Study of Treatment of Renal Insufficiency: Data and Evaluation (STRIDE) (n = 1,348) were combined and analyzed to study the association between serum sodium (Na+) and clinical outcomes. Results: Baseline estimated glomerular filtration rate (eGFR) and Na+ were 26 ± 11 ml/min/1.73 m2 and 140.2 ± 3.4 mEq/l, respectively. The prevalence of Na+ ≤135 mEq/l and ≥144 mEq/l was 6 and 16%, respectively. Higher baseline Na+ was significantly associated with male sex, older age, systolic blood pressure, BMI, serum albumin, presence of heart failure, and lower eGFR. The risk of end-stage renal disease (ESRD) was marginally significantly higher among patients with Na+ ≤135 mEq/l, compared with 140< Na+ <144 mEq/l (referent), in time-dependent models (adjusted hazard ratio, HR = 1.52, p = 0.06). Mortality risk was significantly greater at 135< Na+ ≤140 mEq/l (adjusted HR = 1.68, p = 0.02) and Na+ ≥144 mEq/l (adjusted HR = 2.01, p = 0.01). Conclusion: CKD patients with Na+ ≤135 mEq/l were at a higher risk for progression to ESRD, whereas both lower and higher Na+ levels were associated with a higher risk of mortality. While caring for CKD patients, greater attention to serum sodium levels by clinicians is warranted and could potentially help improve patient outcomes.