Girgis et al ^{Reference Girgis, Phillips, Li, Li, Jiang and Wu1} present data on the usefulness of clozapine versus chlorpromazine in patients with first-episode schizophrenia. The authors must be complimented for conducting a follow-up study of the same cohort after 9 years and being able to have such a high retention rate. Further, the study provides information with respect to the naturalistic setting, reflecting the true clinical situation, and the authors have taken care of possible confounders with appropriate statistical analysis proper explanation. However, there are certain issues with the study. First, the title of the article is somewhat misleading because the randomisation phase of the study was only for the initial 2 years and after that the patients received treatment at the discretion of the clinicians. The title would have been appropriate if the authors were describing the outcome in terms of efficacy/effectiveness and side-effect profile by using survival analysis focusing on either of the medications. But actually the authors describe the effect of clozapine and chlorpromazine for the initial 1 year and outcome at the 9-year follow-up. Second, we need to understand that there are controversies in relation to the definition of first-episode psychosis and the definition used by the authors may appear to be very broad. ^{Reference Breitborde, Srihari and Woods2} Third, the sample size in each treatment group that remained on the same medication (clozapine (n=21) or chlorpromazine (n=8)) at the 9-year follow-up is too small to generalise. Hence, to conclude that there is no difference between clozapine and chlorpromazine with respect to effectiveness would be wrong. Fourth, the authors also conclude that there is no difference in metabolic and other side-effects between the two groups; besides having incomplete baseline data for weight there is no mention of other metabolic variables such as high-density lipoprotein, triglyceride and blood pressure. Fifth, more than half of the study sample (55% of the chlorpromazine group v. 73% of the clozapine group) was not on any anti-psychotic medication at 9-year follow-up, but the authors have not elaborated about their clinical status. Last of all, a quarter of participants (24%) were diagnosed with schizophreniform disorder which might have directly affected the outcome as this group of disorders is considered to have better outcome than schizophrenia.