We investigated whether the disproportionate slowing of schizophrenic patients on cross-modal relative to ipsimodal sequences in a reaction time task ("modality shift effect") could be attributed to a psychometric artifact, as implied by Chapman and Chapman. Fifteen schizophrenic patients (Research Diagnostic Criteria) and 50 normal controls with no current or past history of psychiatric disorder were tested. Subjects made an identical finger-lift response to brief duration light and sound stimuli of different intensities presented in quasi-random order. The major finding was that psychometric artifact could not account for the disproportionate slowing of the reaction time of schizophrenic patients since the reliabilities and variances of the cross-modal and ipsimodal conditions did not differ in normal controls. Furthermore, the modality shift effect was highly significant for reaction times to both sound and light and for all intensities. Findings were the same when the schizophrenic patients were compared to a matched subgroup of slow-responding normal controls.