In addition, we submit that the literature on cardiac performance with vasopressin in patients with septic shock is not entirely as uniform as the authors imply. For example, the authors describe a study by Malay et al,2in which vasopressin exerted no significant effect on cardiac index. Tsuneyoshi et al3 recently reported 16 patients with septic shock treated with continuous vasopressin infusion in whom no significant change in cardiac index was noted. Both of these trials utilized “physiologic” (0.04 U/min) doses of vasopressin, which might theoretically minimize the potential for coronary ischemia, and excluded patients with baseline myocardial dysfunction. While these are small studies, the variable response of cardiac output to vasopressin demonstrated thus far in human trials likely indicates that septic shock is sufficiently heterogeneous that predicting individual responses to such infusions is not straightforward. In particular, while use of vasopressin appears to allow for withdrawal of other vasoactive medications,2,4 in patients with sepsis-associated myocardial depression, addition of inotropes may be required. Randomized, controlled trials will be instrumental in assessing both the efficacy of and specific indications for vasopressin in patients with septic shock.