Purpose: To determine the effects of amifostine on an isolated perfused rat-heart model and its protective activity with regard to cardiotoxic doxorubicin perfusion. Methods: Langendorff constant-pressure isolated rat-heart preparations were used to analyze the effects of the drugs during a 40-min period of perfusion after a 20-min stabilization interval. The first study was conducted with amifostine alone (controls and 10−6, 10−5, and 10−4 M amifostine; n=6 in each group). The second study was conducted with amifostine and doxorubicin (controls, 2.5 × 10−5 M doxorubicin, 2.5 × 10−5 M doxorubicin and 10−5 M amifostine, and 2.5 × 10−5 M doxorubicin and 10−4 M amifostine; n=4 in each group). Results: Amifostine had no significant effect on hemodynamic parameters at 10−6, 10−5, and 10−4 M concentrations. However, amifostine increased the coronary flow expressed as a percentage ± SEM of the baseline flow as follows: 82 ± 4% for controls, 95 ± 6% for 10−6 M amifostine, (P=0.13), 111 ± 4% for 10−5 M amifostine (P < 0.01), and 104 ± 3% for 10−6 M amifostine (P < 0.01). When we commenced an amifostine perfusion 20 min in advance of and then during a 40-min perfusion with doxorubicin, at a cardiotoxic concentration of 2.5 × 10−5 M the left ventricular pressures (LVDP, expressed as percentages ± SEM of the baseline LVDP before doxorubicin) were 55 ± 3% for the doxorubicin controls, 68 ± 2% for doxorubicin with 10−5 M amifostine (P=0.05), and 80 ± 3% for doxorubicin with 10−4 M amifostine (P < 0.01). Whether this protective effect might be related to the known free-radical-scavenging activity of amifostine remains to be determined. Conclusion: On a Langendorff-type model of rat heart, 10−5 and 10−4 M amifostine alone induced a coronary dilation and, when associated with a cardiotoxic concentration of 2.5 × 10−5 M doxorubicin, 10−5 and 10−4 M amifostine displayed a cardioprotective effect.