Even with the availability of potent and selective serotonin antagonists, chemotherapy-induced nausea and vomiting remain a major problem for many patients. This study aims to evaluate the benefit of combination therapy based on Navoban® (tropisetron) in patients who had incomplete control of nausea and/or vomiting induced by chemotherapy when using Navoban® as a single antiemetic agent. In their first chemotherapy course, 1072 patients planned to receive at least two identical cycles of emetogenic chemotherapy were treated with 5 mg Navoban® once daily. To evaluate three treatments additional to the recommended 5 mg once-daily Navoban® regimen during Course 2 in those patients who had shown incomplete control of nausea and/or vomiting on any day of Course 1, a 2×2×2 factorial design was employed. Of these patients, 445 were centrally randomised to receive an additional dose of open-label dexamethasone (Day 1, 0.2 mg/kg i.v.; Days 2–6, 8 mg p.o.) and/or open-label alizapride (Day 1, 100 mg i.v. and 4 × 50 mg p.o.; Days 2–6, 4×5 mg p.o.) and/or double-blind Navoban® - that is, doubling the dose to 10 mg once daily - or placebo. Complete response rates during Course 1 (CRR, no nausea and no vomiting) were, for Day 1, 72% and for Days 1–6, 48%. More complete responders were observed when dexamethasone was added during Course 2, both on Day 1 (76% vs. 66%, p = 0.020) and on Days 1–6 (50% vs. 34%, p = 0.0004). With the addition of a conventional dose of alizapride, a moderate increase in CRR was observed; this was 75% vs. 68% (p = 0.14) on Day 1 and 47% vs. 37% (p = 0.041) on Days 1–6. CRR was not altered when the Navoban® dose was doubled. In patients incompletely controlled with Navoban® alone, the addition of dexamethasone significantly increased the CRR of both acute and delayed emesis (see Discussion).