Background: Somatic mutations in the KCNJ5 potassium channel involve up to 70% of the patients presenting with primary aldosteronism (PA) caused by an aldosterone-producing adenoma (APA). Macrolides corrected the altered function of the two common mutations G151R and L168R, in vitro suggesting their potential usefulness for personalized management of patients with KCNJ5-mutated APA. Hence, we investigated if the macrolide roxithromycin that restored the KCNJ5 channel function could lower plasma aldosterone and blood pressure (BP) in KCNJ5-mutated APA patients. Methods and Design: In a within-patient study we assessed changes of plasma aldosterone, active renin, cortisol and BP induced by an oral roxithromycin (150-300 mg) challenge, in consenting consecutive hypertensive patients. In parallel, we investigated the vascular and BP effects of macrolides ex vivo and in vivo in mice. Results: We recruited 425 consecutive patients: 19 had G151R - or L168R - mutated APA, 27 KCNJ5 wild-type APA, and the rest had no identifiable cause of hypertension. Roxithromycin lowered the plasma aldosterone concentration (p<0.001), not BP, in the G151R- and L168R-mutated APA patients, but not in the wild-type. However, an aldosterone-independent BP decrease was seen in the much larger cohort of hypertensive patients without PA. In mice experiments, studies showed that this was accounted far by a significant anti-hypertensive effect of macrolides that involved an endothelium-NO-dependent vasodilatory action. Conclusions: The roxithromycin-induced fall of plasma aldosterone concentration might be a marker of the presence of KCNJ5-mutated APA. Moreover, our studies unveiled an anti-hypertensive effect of macrolides that involves endothelium- and NO-dependent vasodilation.