Although many trials have evaluated the severity and treatment of nausea and vomiting immediately after cisplatin administration, no studies have focused on vomiting occurring more than 24 hours after chemotherapy--delayed emesis. Two consecutive trials were undertaken to evaluate the incidence, course (trial 1), and severity (trial 2) of delayed nausea and emesis and to develop methods to study these conditions. Eighty-six patients receiving cisplatin (120 mg/m2) for the first time were entered. On the day of cisplatin treatment, all received intravenous (IV) metoclopramide (3 mg/kg X 2 doses) plus dexamethasone (20 mg IV X 1 dose) with either diphenhydramine (50 mg IV) or lorazepam (1.0 to 1.5 mg/m2). Sixty-two percent of patients experienced no vomiting during the 24 hours immediately after administration of cisplatin. Overall, 93% of studied patients experienced some degree of delayed nausea or vomiting from 24 to 120 hours after cisplatin. In trial 1, the incidence of delayed vomiting ranged from 21% to 61% and delayed nausea from 24% to 78% in 58 patients. The highest incidence of both delayed nausea and emesis occurred during the period from 48 to 72 hours after administration of cisplatin. Patients who had no emesis during the initial 24 hours after cisplatin were less likely to experience delayed emesis. The severity of delayed nausea and vomiting was evaluated in 28 patients in trial 2. The amount of delayed nausea and vomiting was assessed daily by patients using a visual analogue scale and by an observer rating. The highest nausea and vomiting scores were seen during the period from 48 to 72 hours after administration of cisplatin, with acceptable correlation between patient scores and observer ratings. Although the nausea and vomiting occurring 24 or more hours after cisplatin administration is not as severe as that seen during the initial 24 hours after administration of cisplatin in patients not receiving antiemetics, it is a common condition that merits both further study and specific treatment.