We were pleased to learn of the recent replication by Stratta et al of our findings with the A-X Continuous Performance Test (CPT), and are grateful that a group of respected colleagues felt this paradigm worthy of further study. Given the number of conflicting findings and methodological complexities present in studies of cognition in schizophrenia, it is critical to demonstrate replications across laboratories as well as within laboratories. As presented in the article by Servan-Schreiber et al,1 we found selective deficits in context processing only among unmedicated patients with schizophrenia. In contrast, Stratta et al found such deficits in medicated patients with multiepisode schizophrenia. Interestingly, we have also recently replicated this finding with medicated patients with multiepisode schizophrenia.2 As pointed out by Stratta et al, such findings raise new questions about the relationship between context processing deficits and medication status, stage of illness, and clinical symptoms. Given the cross-sectional nature of both our own research and that of Stratta et al, future studies using longitudinal designs may be particularly helpful in resolving these questions. By examining the same patients at multiple points (ie, medicated and unmedicated, acute and remitted illness, early and later in illness course), it may be possible to determine the role context processing deficits play in the pathophysiology of schizophrenia.