First, Drs. Wang et al. separately examined dropout rates as a result of "relapse" or "any reason other than relapse." However, overall treatment failure as a result of any reason is a more clinically pragmatic outcome. Had this outcome measure been adopted, the significant superiority in the no dose reduction group would have disappeared (χ2=1.57, df=2, p=0.455). Second, positive effects of reducing risperidone exposure on cognition (2) and subjective experience (3) were not considered. Third, the mean doses in the 4-week and 26-week groups were 2.2 mg/day (SD=0.4) and 2.1 mg/day (SD=0.3), respectively, after the dose reduction. This indicates that a number of subjects were maintained on less than 2.0 mg/day. Given that the lowest limit of dose range of risperidone approved for younger adults with schizophrenia is 2.0 mg/day, reducing the dose below this limit is not standard practice and may raise some ethical considerations. In a recent meta-analysis of published double-blind randomized controlled trials, we demonstrated that moderately low dose treatment may be as effective as the standard dose therapy for relapse prevention, while adopting a very low dose strategy (less than one-half the standard dose) is inferior to both standard and moderately low dose strategies (4). Fourth, a lack of dose retitration phase following symptom worsening limits translation of these results to clinical practice, since transient increase in clinical symptoms would be expected to be reversible with a small dose increase (5). Finally, inspection of Figure 2 and Figure 3 suggests that the 26-week group experienced worsening well before the dose reduction, limiting the interpretation of the observed differences compared with the maintenance group.