Dexrazoxane reduces the cardiotoxicity of anthracyclines, without affecting its antitumor activity. Doxorubicin-induced cardiomyopathy is thought to be due to iron-based oxidative stress. Dexrazoxane is thought to act by hydrolyzing in vivo to ADR-925, a metal ion chelator, which displaces Fe$\sp{3+}$ from doxorubicin. It was demonstrated that ADR-925 and the one-ring open hydrolysis intermediates of dexrazoxane, B and C, are effective chelators, which are able o completely displace Fe$\sp{3+}$ from anthracyclines. Thus, B and C may be pharmacologically active. Titrations of Fe$\sp{3+}$-B with N$\sb3\sp-$ and daunorubicin demonstrated the existence of ternary complexes, and models of Fe$\sp{3+}$-(B)-(H$\sb2$O)$\sb2$ and Fe$\sp{3+}$-(B)-(daunorubicin) complexes were proposed. Fe$\sp{2+}$ and Fe$\sp{3+}$ promoted the hydrolysis of B and C to ADR-925 by factors of up to 6000 and 8, respectively. The pH dependence of these hydrolysis reactions were consistent with hydroxide ion catalysis. Mn$\sp{2+}$, Co$\sp{2+}$, Ni$\sp{2+}$, Cu$\sp{2+}$, and Zn$\sp{2+}$ were also found to promote the hydrolysis of B and C, by factors of 25 to $>$50,000. Physiological concentrations of Mg$\sp{2+}$ and Ca$\sp{2+}$ also promoted the hydrolysis of B and C; these ions may mediate the formation of D, the most effective chelator, in vivo. In addition to promoting the hydrolysis of B and C, Zn$\sp{2+}$ promoted the hydrolysis of dexrazoxane itself. The kinetics of base-catalyzed hydrolysis of a series of imides was characterized. Data from molecular mechanics and semi-empirical calculations were regressed against the kinetic parameters to yield a quantitative structure-activity relationship between imide hydrolysis rates and molecular modelling parameters, which was used to predict the hydrolysis rates of a series of analogs of dexrazoxane. Analogs which were predicted to hydrolyze 2-5 times faster than dexrazoxane, and may therefore be more active, were identified as target molecules.