We thank Roush et al for their letter. 1 We believe both their 2 and our analyses 3 provide some support for recent suggestions that chlorthalidone may be superior to hydrochlorothiazide (HCT) for prevention of cardiovascular events at doses typically used in the clinical setting. 4,5 The data that they show relating relative risk of cardiovascular events to reduction in office systolic blood pressure by HCT and chlorthalidone are interesting, but we note that, for HCT, this regression analysis essentially relies on a single unblinded small study lacking placebo control, 6 with wide confidence limits to the estimated relative risk (0.82 [95% CI, 0.46 -1.44]), where achieved office blood pressures in treatment and control are unavailable, and where only 36% of people received hydrochlorothiazide as monotherapy (the other agents used in combination being propranolol and ␣-methyl-dopa).7 We think the suggestion that chlorthalidone may possess pleotropic actions (or, conversely, that HCT exerts adverse pleotropic actions) is plausible but merits further study. We and Roush et al 1 are in agreement in considering that pharmacokinetic differences resulting in differences in 24-hour blood pressure control are likely to contribute to at least some of the observed differences between HCT and chlorthalidone in terms of cardiovascular disease prevention.