We studied the effects of treatment with olmesartan/amlodipine and olmesartan/hydrochlorothiazide on inflammatory and metabolic parameters (including new-onset diabetes as a secondary endpoint) in non-diabetic hypertensive patients with metabolic syndrome (MetS). A total of 120 patients with MetS and stage I and II hypertension were randomized to olmesartan 20βmg/amlodipine 5βmg or olmesartan 20βmg/hydrochlorothiazide 12.5βmg. If target systolic blood pressure (<140βmmβHg) was not reached, doses were doubled after 13 weeks; doxazosin 4βmg was added after 26 weeks, and doubled after 39 weeks; follow-up ended at 78 weeks. At each visit, blood pressure (BP), fasting plasma glucose, insulin, adiponectin, tumour necrosis factor-Ξ±, C-reactive protein (CRP), intercellular adhesion molecule-1, vascular cell adhesion molecule-1, interleukins-1Ξ², -6 and -8, and albuminuria were measured; BP was similarly reduced in both groups; 80% of patients reached target BP. Reductions in albuminuria were also similar (50%). Only olmesartan/amlodipine reduced the insulin resistance index (24%, P<0.01), increased plasma adiponectin (16%, P<0.05) and significantly reduced all of the inflammation markers studied, except CRP, which showed a similar reduction in each group. The risk of new-onset diabetes was significantly lower with olmesartan/amlodipine (P=0.02). Both olmesartan-based combinations were effective, but the amlodipine combination resulted in metabolic and anti-inflammatory effects that may have advantages over the hydrochlorothiazide combination.