Hypertension-induced structural remodeling of the left atrium (LA) has been suggested to involve the renin-angiotensin system. This study investigated whether treatment with an angiotensin receptor blocker, candesartan, regresses atrial remodeling in spontaneously hypertensive rats (SHR). Effects of treatment with candesartan were compared to treatment with a nonspecific vasodilatator, hydralazine. Thirty to 32-week-old adult male SHR were either untreated (n = 15) or received one of either candesartan cilexetil (n = 9; 3 mg/kg/day) or hydralazine (n = 10; 14 mg/kg/day) via their drinking water for 14 weeks prior to experiments. Untreated age- and sex-matched Wistar-Kyoto rats (WKY; n = 13) represented a normotensive control group. Untreated SHR were hypertensive, with left ventricular hypertrophy (LVH) compared to WKY, but there were no differences in systolic pressures in excised, perfused hearts. LA from SHR were hypertrophied and showed increased fibrosis compared to those from WKY, but there was no change in connexin-43 expression or phosphorylation. Treatment with candesartan reduced systolic tail artery pressures of conscious SHR below those of normotensive WKY and caused regression of both LVH and LA hypertrophy. Although hydralazine reduced SHR arterial pressures to those of WKY and led to regression of LA hypertrophy, it had no significant effect on LVH. Notably, LA fibrosis was unaffected by treatment with either agent. These data show that candesartan, at a dose sufficient to reduce blood pressure and LVH, did not cause regression of LA fibrosis in hypertensive rats. On the other hand, the data also suggest that normalization of arterial pressure can lead to the regression of LA hypertrophy.