Abstract Background and Aims Hyperkalemia (HK) is a common and dangerous complication of CKD because of impaired kidneýs ability for potassium elimination. On the other hand, HK is a common complication of extremely beneficial therapeutic agents acting on the renin–angiotensin–aldosterone system (RAAS). Its initiation at early CKD stages is even more benefic but HK could lead to stop it. We wonder if there is a possible relation between HK, therapeutic changes in RAAS inhibition (not initiating or stopping it) and mortality. Our goal was to investigate incidence, prevalence and clinical outcomes of at least one episode of HK in a CKD population outpatient setting. Additionally, we investigated the association of HK with changes in RAAS inhibition and mortality risk. Method We conducted a patient-level, retrospective, cohort analysis of all adult patients referred to a nephrology clinic over a 6 years period. We included CKD stage 3 patients with at least 24 months of follow up and three or more serum potassium determinations. The prevalence of HK (blood potassium level ≥ 5,5mmol/L) at first consultation and incidence during follow up were accessed. Patients were spited in two groups prior to analysis: A) Patients without any HK episode and B) Patients with at least one HK episode. Baseline and follow up covariates included demographics, comorbid conditions, laboratory values, HK-associated drugs [ACEis, ARBs, potassium-sparing diuretics and diuretics]. The impact of HK and therapeutic changes on mortality was evaluated through a logistic regression. Results Out of the 3008 patients referred to the nephrology clinic, 575 (19.1%) met the inclusion criteria (mean age: 70.4 years; 63.7% male and 94.0% caucasians). Mean follow-up was 4.1±1.8 years. Important cardiovascular comorbidities included hypertension (HTN) (90.3%); overweigh (67.4%), DM (49.0%) and Heart Failure (31.4%). CKD stage progression was present in 122 (21.2%). The prevalence of HK at first consultation was 8.7% and follow up incidence 21.7%. From this cohort, 164 (28.5%) had at least on episode of HK (Group B) and 101 (17.6%) died. During the follow up, RAAS inhibition drugs was removed or not started in 200 (34.8%) patients and diuretic was initiated in 165 (28.7%). In univariate analysis, at least one HK episode was associated with Diabetes (65.9 vs 42.3%, p<0.001), Heart failure (36.6 vs 28.0%, p=0.007), Macroalbuminuria (34.1 vs 21.2%, p=0.001), CKD progression (33.5 vs 16.3. p<0.001) higher frequency of diuretic initiation (38.4 vs 24.8%, p<0.001) and higher mortality (27.6 vs 13.7%, p<0.001). In multivariate logistic regression analysis, the independent predictors of mortality were: At least one HK episode (OR 1.82, 95% CI 1.08-3.04, p=0.02); Heart Failure (OR 1.97, 95% CI 1.16-3.35, p=0.01); Older age (OR per 1 year increase 1.04, 95% CI 1.02-1.07, p=0.001); CKD progression (OR 4.18, 95% CI 2.43-7.19, p<0.001). Predictors of lower mortality risk were: Patients who maintained RAAS inhibition during follow up (OR 0.50, 95% CI 0.26-0.96, p=0.03); Patients who started RAAS inhibition during follow up (OR 0.38, 95% CI 0.16-0.88, p=0.02). Conclusion Our study confirms that RAAS inhibition had a protector and independent impact in mortality when prescribed in CKD early stages. On the other hand, patients with at least one episode of HK have a higher risk of mortality. All efforts should be made to maintain these therapeutic agents, looking for other ways to control hyperkalemia rather than stop it.