A phase I trial and pharmacokinetic study of 5-aza-2'-deoxycytidine (5-aza-dCyd) were conducted in 21 patients with advanced solid tumors. The drug was given as three 1-h infusions, separated by intervals of 7 h. Treatment was repeated every 3-6 weeks. Forty-six cycles of 5-aza-dCyd were administered at 7 dose levels ranging from 25 to 100 mg/m2 in three infusions. The dose-limiting toxicity was myelosuppression, with a delayed white blood cell nadir, occurring at Day 22. Other toxicities included a mild, reversible elevation of serum creatinine in three patients, minimal nausea and vomiting in six patients, and transient fatigue in three patients. In this study one partial response in a patient with an undifferentiated carcinoma of the ethmoid sinus was observed. Plasma and urinary concentrations of 5-aza-dCyd were measured using a bioassay based on growth inhibition of L1210 leukemia cells in vitro. For 75 and 100 mg/m2 given as 1-h infusions, mean peak plasma concentrations of 0.93 and 2.01 microM, respectively, were attained. In seven of nine courses at doses of 25-60 mg/m2, plasma 5-aza-dCyd concentration was less than 0.01 microM. In one case at 30 mg/m2 and another at 60 mg/m2, peak plasma drug concentrations were determined to be 0.244 and 0.409 microM, respectively. Following cessation of the infusion rapid disappearance of drug from plasma was observed with a t1/2 alpha and t1/2 beta of 7 and 35 min, respectively. High clearance values and a total urinary excretion of less than 1% of the administered dose suggest that 5-aza-dCyd is eliminated rapidly and largely by metabolic processes. For the present schedule studied, a dose of 75 mg/m2 in three infusions, every 5 weeks, is recommended for phase II trials in solid tumors.