The combination of carboplatin and paclitaxel (Taxol; Bristol-Myers Squibb Company, Princeton, NJ) has been found to be highly active, to offer convenience of administration and better potential for dose escalation, and to produce less nonhematologic toxicity than standard therapy with cisplatin plus paclitaxel. Increased myelosuppression was anticipated to be a disadvantage of the carboplatin/paclitaxel combination, although it was expected to affect mainly platelets. Studies to date of the paclitaxel/carboplatin combination suggest the combination has a relatively benign effect on platelets compared with single-agent carboplatin. Previous pharmacokinetic studies have suggested that the measured area under the plasma carboplatin concentration-time curve is significantly less than that predicted by dosing formulas based on glomerular filtration rates. Newer evidence, however, contradicts these findings. This review examines the evidence of both a pharmacodynamic and a pharmacokinetic interaction between carboplatin and paclitaxel that has a sparing effect on platelets. It further assesses methodologic differences in methods used to estimate glomerular filtration. Paclitaxel appears to offer some protective effect for carboplatin-induced thrombocytopenia, although the two drugs do not interact pharmacokinetically. The apparent deviations of the achieved carboplatin area under the plasma concentration-time curve from that predicted by a dosing formula are thought to be due to differences in the methodology used to estimate renal function and measure creatinine. Any interaction between the two drugs most likely occurs at the level of the megakaryocyte.