INTRODUCTION: The current debate about the side effects of induction agents, e.g. possible adrenal suppression through etomidate, emphasizes the relevance of choosing the correct induction agent in septic patients. However, cardiovascular depression is still the most prominent adverse effect of these agents, and might be especially hazardous in septic patients presenting with a biventricular cardiac dysfunction--or so-called septic cardiomyopathy. Therefore, we tested the dose-response direct cardiac effects of clinically available induction agents in an isolated septic rat heart model. METHODS: A polymicrobial sepsis was induced via cecal ligation and single puncture. Hearts (n = 50) were isolated and randomly assigned to five groups, each receiving etomidate, s(+)-ketamine, midazolam, propofol, or methohexitone at concentrations of 1 x 10-8 to 1 x 10-4 M. Left ventricular pressure, contractility and lusitropy, and coronary flow were measured. Cardiac work, myocardial oxygen delivery, oxygen consumption, and percentage of oxygen extraction were calculated. RESULTS: All of the induction agents tested showed a dose-dependent depression of cardiac work. Maximal cardiac work dysfunction occurred in the rank order of s(+)-ketamine (-6%) <etomidate (-17%) <methohexitone (-31%) <midazolam (-38%) <propofol (-50%). In addition, propofol showed a maximum decrease in contractility of -38%, a reduction in lusitropy of -44%, and a direct vasodilator effect by increasing coronary flow by +29%. CONCLUSIONS: Overall, this study demonstrates that these tested drugs indeed have differential direct cardiac effects in the isolated septic heart. Propofol showed the most pronounced adverse direct cardiac effects. In contrast, S(+)ketamine showed cardiovascular stability over a wide range of concentrations, and might therefore be a beneficial alternative to etomidate.