Histamine-2 receptor antagonists (H2RAs) are principal components of the premedication regimen used to prevent major hypersensitivity reactions in patients receiving paclitaxel. Several different H2RAs, including cimetidine, ranitidine and famotidine, have been used in clinical trials of paclitaxel, as well as by clinicians in different geographic regions and hospitals primarily because of differences in the availability of the various H2RAs. However, H2RAs have highly variable cytochrome P450-modulating capabilities, and the P450 system appears to play a major role in paclitaxel metabolism and disposition. Therefore, the use of different H2RAs may result in different pharmacologic, toxicologic and antitumor profiles due to differential effects on paclitaxel metabolism. This study evaluated whether cimetidine and famotidine, which possess disparate P450-modulating capabilities, differentially affect paclitaxel clearance rates and the agent's principal toxicity, neutropenia. Women with advanced, platinum-refractory ovarian carcinoma received two courses of treatment with 135 mg/m2 paclitaxel over 24 h while participating in the National Cancer Institute's Treatment Referral Center Protocol. A crossover design was employed in which consecutive patients received either 300 mg cimetidine i.v. or 20 mg famotidine i.v. before their first course of paclitaxel and the alternate H2RA before their second course. In order to evaluate the differential effects of cimetidine and famotidine on pertinent pharmacologic and toxicologic parameters in the same individual, paclitaxel concentrations at steady-state (Css), paclitaxel clearance rates, and absolute neutrophil counts (ANCs) were obtained during both courses. Paclitaxel Css values were not significantly different in individual patients when either cimetidine or famotidine preceded paclitaxel (p = 0.16). Mean paclitaxel clearance rates were 271 and 243 ml/min per m2 following cimetidine and famotidine, respectively. These clearance rates were not significantly different in paired analysis (p = 0.30). The likelihood of subsequently requiring granulocyte-colony stimulating factor (G-CSF) for severe neutropenia during course 1 did not differ significantly between the two H2RAs (p = 0.9). Among patients who did not require G-CSF, mean percentage decreases in ANC were 87.7% and 84.2% after paclitaxel cycles preceded by cimetidine and famotidine, respectively. These measures of neutropenia did not differ significantly in paired analysis (p = 0.13). These results show that the H2RAs cimetidine and famotidine do not differentially affect the pharmacologic and toxicity profiles of paclitaxel when used in the premediation regimen to prevent major hypersensitivity reactions, and may be interchanged.