OBJECTIVE We performed a dose-escalation study of 2-chlorodeoxyadenosine (2-CdA) in solid tumors to determine the maximum-tolerated dose (MTD) and define its toxicity profile at higher doses. METHODS Twenty-one patients, seven with malignant astrocytoma, twelve with metastatic melanoma, and two with metastatic hypernephroma, were enrolled onto the study. Patients were entered onto cohorts that received 0.10, 0.15, or 0.20 mg/kg/d of 2-CdA by continuous intravenous infusion for 7 days every 28 days. 2-CdA levels were determined by radioimmunoassay. In tumor tissue samples, deoxycytidine kinase (dCK) levels were measured by both enzyme activity and immunoreactive protein analysis. RESULTS Of seven patients treated with 2-CdA at 0.1 mg/kg/d, one experienced grade 3 or 4 myelotoxicity. Of 11 patients treated at 0.15 mg/kg/d, four experienced myelotoxicity, two after a single course of 2-CdA. All three patients who received 2-CdA at 0.2 mg/kg/d experienced myelosuppression. Neurologic events occurred in two patients, both with malignant melanoma. Two of seven patients (28.6%) with astrocytomas obtained partial responses with a median duration of 8 months. 2-CdA penetrated the blood-brain barrier. An association was found between dCK levels as measured by enzymatic activity and immunoreactive proteins, but this did not correlate with 2-CdA tumor responsiveness. CONCLUSIONS The MTD for 2-CdA delivered as a 7-day intravenous infusion in patients with nonhematologic malignancies was determined to be 0.1 mg/kg/d, the same as the MTD for patients with hematologic malignancies. There was no clinical correlation with dCK expression and response to 2-CdA. The responses noted in patients with malignant astrocytoma warrant further phase II study.