Trimetrexate (TMQ; NSC 352122) is a potent inhibitor of dihydrofolate reductase with good activity against murine i.p.-implanted B16 melanoma and colon 26 tumors. Preclinical antineoplastic activity, demonstrated schedule dependency, and data suggesting effectiveness against methotrexate-resistant cells prompted a Phase I clinical and pharmacokinetic study of trimetrexate using an i.v. daily x5 schedule. Forty-three good performance status patients were treated with 12 dose levels using daily doses varying from 0.5 to 15 mg/m2/d. Plasma and urine samples were obtained for pharmacokinetic analysis using a high-performance liquid chromatographic method. Myelosuppression was dose limiting and 15 mg/m2/d x5 was the maximum tolerated dose. White blood cell (WBC) and platelet toxicity were noted at doses of 1.6 mg/m2 and above. Median WBC and platelet nadirs occurred on approximately Days 11-12 with recovery by Days 15-18. Nonhematological toxicity included mucositis, nausea and vomiting, stomatitis, diarrhea, and rash. Evidence for antitumor activity was seen in seven patients. Trimetrexate elimination from plasma could be represented as either a bi- or triexponential process. Terminal elimination half-lives were in the range of 5-14 h in patients represented by a triexponential model. Approximately 10-20% of the dose administered was excreted in urine over a 24-h period. The recommended starting dose for patients in Phase II trials using the d x5 i.v. schedule is 8.0 mg/m2/d repeated every 21 days. Dose escalations may be possible depending on the extent of prior therapy and individual tolerance of the drug.