The aim of the present randomized study was to evaluate which dose of Ondansentron (OND)(32 versus 8 mg) is appropriate for the antiemetic treatment of a uniform group of patients (pts) with Non Small Cell Lung Cancer (NSCLC) who were treated with Cisplatin (CDDP) 100 mg/m2 in combination with other less emetogenic drugs. One hundred and ten patients, with histologically confirmed NSCLC entered this randomized study. They were between 50 - 70 years old, with no previous Chemotherapy, with a PS (Karnofsky) >60%. They were randomized into two groups; Group A: OND as a 32 mg dose the first 24 hours, followed by 8 mg every 8 hrs for the following four days, combined with dexamethasone, 8 mg i.v. the first day, and 8 mg p.o., in the morning, the following three days. Group B: OND as a 8 mg dose every day for 4 days, combined with dexamethasone 8 mg i.v. and 8 mg p.o. the following three days. In this randomized study, of the 110 patients who entered, 106 were evaluable. Clinical parameters were similar between the examined groups. A higher number of patients of Group A presented complete response (P 0.0001), compared to patients of Group B who failed (P 0.004), during the first 24 hours. In the 3 days that followed, a higher number of pts of Group A presented complete response to the antiemetic therapy (P 0.001, P 0.0001), while Group B failed (P 0.007, P 0.001, P 0.019), or presented minor response (P 0.0001, P 0.004). Patients who had no antiemetic response needed additional therapy and were excluded from the evaluatio (13 pts of Group B). Retches (P 0.0001, P 0.005), and nausea (P 0.0001, P were also frequent in Group B. We concluded that reduced OND doses (8 mg) are inadequate in the prevention of emesis after high dose CDDP (100 mg/m2) and should be avoided.