A double-blind randomised controlled trial comparing the antiemetic effects of sublingual lorazepam combined with high-dose, short course metoclopramide (3 mg/kg) infused twice 3 h apart with or without i.v. dexamethasone is reported. Sixty patients receiving a total of 209 cycles of potentially severely emetic cytotoxic chemotherapy were randomised to receive one or other antiemetic regimen. In those receiving platinum-based chemotherapy the addition of dexamethasone was associated with an improvement in freedom from nausea (P less than 0.01) and freedom from vomiting (P less than 0.05). In the non-platinum-based chemotherapy group the addition of dexamethasone led to a reduction in the duration and severity of nausea, and duration of vomiting (P less than 0.05 in each case). Both antiemetic regimens were well tolerated with a low incidence of adverse effects and could be administered easily in an outpatient setting.