Chemotherapy-induced nausea and vomiting can today be controlled with the "new" antiemetics or with their various combinations in a high percentage of patients. Despite this, for some subgroups of patients, certain chemotherapy regimens and some aspects of the phenomenon (delayed presentation), emesis remains a critical problem. Hence, there is the necessity for a continuous effort in the search for new drugs or better treatment modalities and the absolute requisite that these efforts be carried out according to a sound and verified trial methodology. Nausea and vomiting induced by antineoplastic agents are extremely variable phenomena, depending not only on the characteristics of chemotherapy regimens and of the patient population, but also on a subjective feeling generated by the impact of the care system on the patient's individual situation. Therefore, precisely because of this high variability, large comparative trials should be carried out to ensure that the sample is sufficiently representative for the most efficacious antiemetic regimen to be detected. In this field, some of the main problems arising in all clinical trials have their own specificity, particularly the study design, whether completely randomized or cross-over, the follow-up and the importance of prognostic factors. Among these, age, gender and previous chemotherapy with experience of nausea and vomiting have been confirmed as important. In addition, some topics must clearly be highlighted: the definition of the response variables and the assessment of variability of the results obtained with the same antiemetic regimen from one cycle of chemotherapy to the next (i.e. persistence). The greater attention devoted today to delayed emesis raises some methodological questions: this paper suggests some possible solutions for a better evaluation of this phenomenon.