Different cytotoxic drugs induce different patterns of emesis. This is relevant to clinical practice since we often see in the medical literature oversimplifications in the recommendation for management of chemotherapy-induced emesis, so that the same guidelines are given for cisplatin and non-cisplatin-containing chemotherapy. In particular, cisplatin induces a biphasic pattern of emesis which is characterized by an acute immediate phase and a delayed phase. These two phases are clearly different, especially when cisplatin is given in short i.v. administration (e.g. over 20 min to 1 h) and in high doses (100–120 mg/m2). On the other hand, cyclophosphamide and carboplatin induce a quite different pattern of emesis, characterized by a monophasic curve which, although more intense in the first 24 h, may continue for a number of days. The antiemetic response to 5-HT3 receptor antagonists on days 2–5 is good in the case of carboplatin and cyclophosphamide and poor in the case of cisplatin (delayed emesis after cisplatin). This firmly suggests that the pathogenesis of cisplatin-induced emesis may differ from that induced by other cytotoxic drugs, especially in the delayed phase of emesis. We think that we should reserve the term ‘delayed emesis’ for the late emesis induced by cisplatin, while ‘prolonged emesis’ could be a better denomination for the late emesis induced by cyclophosphamide and carboplatin. The main clinical implication of these observations is that 5-HT3 receptor antagonists should be administered over 3–5 days in the case of carboplatin and cyclophosphamide, while a short treatment during the first day could be sufficient to control the acute phase of cisplatin-induced emesis.