In a double-blind, randomised, multicentre study, the efficacy and tolerability of tropisetron and a combination of tropisetron and dexamethasone were compared for the control of nausea and vomiting induced by cisplatin in patients previously not entirely protected by tropisetron monotherapy. In all, 160 women with gynaecological cancers were studied during two consecutive courses of cisplatin-containing chemotherapy. During the first course (the screening course), all patients received tropisetron monotherapy [5 mg intravenous (i.v.) on day 1 and 5 mg orally on days 2-6] as antiemetic treatment. During the second course (the test course), tropisetron was compared with a combination of tropisetron and dexamethasone (20 mg i.v. on day 1 and 4.5 mg twice daily on days 2-6). This part of the study was double-blind, randomised and placebo-controlled. Candidates for randomisation were patients with partial control of nausea (< 12 h of nausea) or partial control of vomiting (1-4 episodes of vomiting) during the screening course. Patients with complete control of nausea and vomiting in the screening course continued with tropisetron monotherapy; patients with treatment failure received open rescue treatment in course 2. Total control of acute nausea was achieved in 37% of the tropisetron + placebo group and in 75% of the tropisetron + dexamethasone group (P = 0.001). Significantly more patients on tropisetron-dexamethasone than on tropisetron-placebo were also free of delayed nausea. Acute vomiting was prevented in 40% of the patients in the placebo group and in 75% in the dexamethasone group (P = 0.001). Delayed vomiting was also significantly less frequent in dexamethasone-treated patients than in placebo-treated patients. Tropisetron was well tolerated both as monotherapy and in combination with dexamethasone. The most frequent adverse events were headache (34%), constipation (12.5%) and fatigue (12.5%). Adding high doses of a corticosteroid did not induce further adverse events or disregulate concurrent diseases.