A phase I study of a 96 h paclitaxel infusion with filgrastim support was performed to determine the toxicity, maximum-tolerated dose (MTD) and pharmacokinetics in patients with refractory solid tumors. In this phase I trial, the initial paclitaxel dose was 140 mg/m2/96 h followed by filgrastim (5 microg/kg/day s.c.) beginning 24 h after the paclitaxel and continued until granulocyte recovery. Cycles were repeated every 21 days. Patients with refractory solid tumors were eligible; however, only one previous chemotherapy regimen was allowed. The dose of paclitaxel was escalated by 20 mg/m2/96 h in subsequent cohorts until dose-limiting toxicity (DLT) occurred. Pharmacokinetic analysis was performed by quantitating paclitaxel concentrations at baseline, 24, 48, 72 and 96 h after the start of the paclitaxel infusion. Twenty-one patients were entered into this trial of which 19 were evaluable. A total of 52 treatment cycles were administered. DLT was seen in two of four patients at 200 mg/m2/96 h, and consisted of diarrhea, mucositis and granulocytopenic infection. The MTD of the 96 h paclitaxel infusion was 180 mg/m2 with filgrastim support. Mucosal and granulocyte toxicity were correlated with steady-state paclitaxel concentrations (Css) greater than 0.100 micromol/l. In the presence of liver function test 1.5 times or lower than normal, metastatic liver disease did not alter paclitaxel Css. Objective responses were observed in non-small cell lung cancer, small cell lung cancer and melanoma. The recommended phase II dose of paclitaxel infused over 96 h with filgrastim support is 180 mg/m2. Paclitaxel Css correlate with mucosal and granulocyte toxicity. In the presence of normal enzymatic function, metastatic liver disease does not affect paclitaxel clearance.