We read with much interest the thorough review article by Kizer and Kimmel1 on the epidemiologic evidence on calcium antagonists. In our opinion, the paragraph addressing the "major hemorrhage" issue might convey a misguided message to the reader. In particular, our study2 on the in-hospital variations in hemoglobin levels was not focused on major hemorrhage, as the outcome variable was a decrease in the hemoglobin level of greater than 1.2 g/dL. Noticeably, that study found an independent association between use of calcium antagonists and the incident decrease in hemoglobin levels among patients with a definite diagnosis and/or treatment for peptic disease (odds ratio [OR] = 1.67; 95% confidence interval [CI] = 1.26-2.22), but not among other participants (OR = 1.02; 95% CI = 0.82-1.25). These figures correspond (as Table 7 in the article by Kizer and Kimmel should have depicted adjusted relative risks [RRs]) to RRs of 1.50 (95% CI = 1.21-1.84) and 1.01 (0.85-1.20), respectively.3 Thus, calcium antagonists do not seem to cause bleeding; rather, their effects on hemostasis might become clinically relevant in subjects with bleeding lesions, such as surgical wounds or peptic erosions.2,4 This finding is consistent with the inhibitory effects of calcium antagonists on in vitro and in vivo platelet aggregation and secretion, which have been documented by several studies.5- 8 In this setting, calcium antagonists should not be considered more "unsafe" than any other antiplatelet agent. Indeed, use of calcium antagonists as adjuvant antiplatelet agents has been proposed by several authors.5- 8 In a more general view, the ongoing debate for or against calcium antagonists should give way to an objective analysis of the impact of these agents in different clinical settings. The review by Kizer and Kimmel represents a substantial advancement toward this goal.