Exercise training (ET) is a coadjuvant therapy in preventive cardiology. It delays cardiac dysfunction and exercise intolerance in heart failure (HF); however, the molecular mechanisms underlying its cardioprotection are poorly understood. We tested the hypothesis that ET would prevent Ca(2+) handling abnormalities and ventricular dysfunction in sympathetic hyperactivity-induced HF mice. A cohort of male wild-type (WT) and congenic alpha(2A)/alpha(2C)-adrenoceptor knockout (alpha(2A)/alpha(2C)ARKO) mice with C57BL6/J genetic background (3-5 mo of age) were randomly assigned into untrained and exercise-trained groups. ET consisted of 8-wk swimming session, 60 min, 5 days/wk. Fractional shortening (FS) was assessed by two-dimensional guided M-mode echocardiography. The protein expression of ryanodine receptor (RyR), phospho-Ser(2809)-RyR, sarcoplasmic reticulum Ca(2+) ATPase (SERCA2), Na(+)/Ca(2+) exchanger (NCX), phospholamban (PLN), phospho-Ser(16)-PLN, and phospho-Thr(17)-PLN were analyzed by Western blotting. At 3 mo of age, no significant difference in FS and exercise tolerance was observed between WT and alpha(2A)/alpha(2C)ARKO mice. At 5 mo, when cardiac dysfunction is associated with lung edema and increased plasma norepinephrine levels, alpha(2A)/alpha(2C)ARKO mice presented reduced FS paralleled by decreased SERCA2 (26%) and NCX (34%). Conversely, alpha(2A)/alpha(2C)ARKO mice displayed increased phospho-Ser(16)-PLN (76%) and phospho-Ser(2809)-RyR (49%). ET in alpha(2A)/alpha(2C)ARKO mice prevented exercise intolerance, ventricular dysfunction, and decreased plasma norepinephrine. ET significantly increased the expression of SERCA2 (58%) and phospho-Ser(16)-PLN (30%) while it restored the expression of phospho-Ser(2809)-RyR to WT levels. Collectively, we provide evidence that improved net balance of Ca(2+) handling proteins paralleled by a decreased sympathetic activity on ET are, at least in part, compensatory mechanisms against deteriorating ventricular function in HF.