A majority of studies show that schizophrenics perform poorly on so-called tests of executive or frontal lobe function--the paradigmatic case being the Wisconsin Card Sort Test (WCST). Nevertheless, the specific character of this deficit in schizophrenia remains underspecified. In particular, it seems premature to assume that schizophrenia is characterised by an executive dysfunction and/or a disorder of frontal lobe function before determining whether any deficit is: selective; disproportionate to the general level of intellectual functioning; or qualitatively comparable with that of frontal lobe patients. A meta-analysis was conducted on 29 studies comparing the performance of schizophrenics and normal controls on the WCST. This showed that the mean weighted effect size was large for categories achieved (d = 0.91), medium for absolute level of perseveration (d = 0.53), but only small for the proportion of perseverative errors (d = 0.18). By contrast, the effect size for Wechsler Adult Intelligence Scale Intelligence Quotient (WAIS IQ) in a subset of these studies (d = 1.23) was significantly larger than for any WCST measures. This pattern of findings challenges notions that schizophrenia is characterised by an executive dysfunction that is: selective; disproportionate to IQ level; and analogous to that found in frontal lobe patients. Rather, the poor WCST performance of schizophrenics appears to reflect a generalised intellectual deficit.