BACKGROUND: Brain morphology and physiological measures in schizophrenia have yielded inconsistent results. This may be due in part to difficulties in ascertaining precisely to what degree each measure deviates from its genetically and environmentally determined potential level. We attempted to surmount this problem in a paradigm involving monozygotic twin pair discordant for schizophrenia. In this paradigm, the difference score and reaction time between the unaffected member and affected member of a twin pair should represent the degree of pathologic involvement irrespective of actual level. METHOD: We investigated, using fMRI, the neural substrate underlying encoding and retrieval of aversive and neutral IAPS pictures. RESULTS: An ANOVA on reaction time (RT) between schizophrenia patient (J) and normal sister (D) significant difference, (F = 5.2, p < or = .02) for J had less RT than D. Conversely, the ANOVA for the correct pictures retrieved was insignificant (F = 1.8, p < or = .2). When the brain activity associated with the encoding and retrieval of the aversive pictures was subtracted (J-D and D-J) from that associated with the neutral ones, significant loci of activation were found. During encoding: for J-D the right fusiform gyrus was significantly activated (p < .0001) and for D-J the orbitofrontal cortex was significantly activated (p < .05). During retrieval: for J-D the right anterior cingulate (p < .0001) was activated and the dorsolateral prefrontal cortex (p < .002). For D-J only the cerebellum showed activation (p < .0001). CONCLUSION: Results indicated subtle attenuations in some aspects of memory, thus providing another evidence for cognitive markers of a genetic component in schizophrenia. New approaches in neuropsychiatry-based on genetic methodologies should further define the cerebral physiology responsible for schizophrenia.