In a multicenter phase II Italian trial that used a 28-day dosing schedule of gemcitabine on days 1, 8, and 15 and cisplatin on day 2, thrombocytopenia and neutropenia were the main dose-limiting toxicities observed. The aim of the present study was to determine whether using 15-day cisplatin in lieu of the standard 2-day schedule in combination with weekly gemcitabine would decrease expected myelotoxicities, particularly thrombocytopenia. Fifty-one patients with advanced non-small cell lung cancer (NSCLC), a median age of 62 years (range 31-76) and baseline Eastern Cooperative Oncology Group (ECOG) performance status scores of 0-1, were enrolled. Twenty-four patients had stage IIIA-B disease and 27 had stage IV. Patients received gemcitabine 1000 mg/m(2) on days 1, 8, 15, and cisplatin 100 mg/m(2) on day-15, every 28 days for a total of 151 cycles. All patients were evaluable for toxicity. Grades 3 and 4 thrombocytopenia was observed in 16% of patients, grades 3 and 4 neutropenia in 35% of patients, and grade 3 anemia in 4% of patients (no grade 4 anemia). Nonhematologic toxicity was mild. Two patients had grade 3 vomiting, and another had grade 4 hepatic toxicity only after gemcitabine administration. The dose intensity of gemcitabine and cisplatin was well maintained. Of the 45 patients evaluable for response, there were 22 (49%) partial responders, 7 (15.5%) minimal responders, 9 (20%) with stable disease, and 7 (15.5%) progressions. Compared with the schedule used in a multicenter phase II Italian trial (day 2 cisplatin), day-15 cisplatin decreases incidences of thrombocytopenia (16 vs. 52%) and anemia (4 vs. 25%); the occurrence of neutropenia is similar (35 vs. 36%). Response rates are also similar (49 vs. 54%).