Functional neuroimaging findings of "hypofrontality" in schizophrenic patients - as tested with the Wisconsin Card Sorting Test (WCST) - are still controversial, mainly due to methodological aspects and the heterogeneity of the patient samples. To measure WCST specific and reproducible reduced cerebral activations in schizophrenic patients, we revised the study design and patient recruitment, respectively. For this purpose, we used an adequate active control task instead of an undefined rest condition to determine exclusively WCST specific cerebral activations. In addition, we focused on the investigation of modified activations between a selected group of neuroleptic-naive schizophrenic patients and carefully matched healthy controls by means of functional magnetic resonance imaging. The results indicate that neuroleptic-naive schizophrenic patients show reduced activations in the right frontal and left temporal lobe, as well as in the left cerebellum. By utilizing an active control task all unwanted activations are suppressed. Furthermore the influence of different task performances is reduced. The findings are in line with previous PET and SPECT studies and confirm the "hypofrontality" hypothesis. The findings suggest that "hypofrontality" is not caused by neuroleptic medication.