dealing with the chemical stability of the novel prodrug antitumor agent carzelesin (U-80,244). In summary, we presented kinetic data and a pH-rate profile and confirmed the identities of some decomposition products of carzelesin. HPLC chromatograms of partly degraded samples showed U-76,073, U-76,074, and aniline as degradation products. On the basis of these results we proposed a degradation scheme of carzelesin in alkaline medium. This involves a nucleophilic attack of an hydroxyl ion on the carbonyl of the carbanilic moiety in carzelesin, followed by cleavage of carbanilate and the formation of U-76,073. Consecutive intramolecular rearrangement in U-76,073 leads to U-76,074; carbanilate is unstable and decomposes to aniline. The presence of peaks in the chromatograms corresponding to U-76,073, U-76,074, and aniline supports this. Dr Bergon, however, questions the proposed mechanism based on reported experiences with hydrolysis of carbamates. He states that an eliminationaddition mechanism through the formation of phenyl isocyanate is the most likely one for carzelesin degradation. However, under our experimental conditions, phenyl isocyanate could not be detected in the decomposition mixtures. This does not appear to be due to the chemical instability of phenyl isocyanate, because it is stable during at least several half-lives of carzelesin under the experimental conditions (e.g., 0.03 M phosphate buffer/acetonitrile, pH 7.2, 25°C) used. In our HPLC system with a mobile phase composed of 2 mM phosphate buffer, pH 6.5/acetonitrile