Both early (i.e., pre- and perinatal periods) and late (i.e., adolescent period) neurodevelopmental processes are thought to participate in the etiology and pathophysiology of schizophrenia. However, whether markers of these processes would be expected to predict an imminent onset of psychosis, as is hoped in the current generation of prodromal research programs, depends on whether their disruptions result from genetic factors shared by patients and some of their unaffected relatives, nongenetic factors specific to those who manifest the illness phenotype, or combinations of these sets of influences. Here we present recent work deriving primarily from high-risk and family-study (i.e., "genetic high-risk") designs, which provide a framework for investigating the neural changes that may occur proximally to the initial onset of psychosis. This work indicates that some of the alterations in brain function and structure in schizophrenia are primarily genetically mediated and also appear in some of patients' unaffected first degree relatives, while other alterations are present in individuals who manifest the illness phenotype but not in relatives at genetic risk (Cannon et al. 2002a, 2002c; Van Erp et al. 2002). Whereas the primarily genetically mediated deficits shared by at-risk but nonsymptomatic relatives are not likely to show differential change in the premorbid period and may be necessary but clearly not sufficient for the development of psychotic symptoms, the deficits specific to patients who manifest the illness phenotype are good candidates for marking the neurobiological processes associated with the emergence of psychotic symptoms at the time of schizophrenia onset. Preliminary results from longitudinal studies of individuals ascertained initially in a prodromal (i.e., "clinical high-risk") state appear to be interpretable within this framework. A number of questions arising from this line of inquiry need to be addressed in the current generation of prodromal research programs: To what extent do the neural systems affected by early and late neurodevelopmental influences overlap? Is there likely to be a schizophrenia-related disturbance in the processes associated with adolescent brain maturation, or are these maturational processes themselves intact, participating in psychosis onset only indirectly, by promoting a neurobiological context in which the early neurodevelopmental disturbances can be expressed in psychotic symptoms? What pattern of changes observable from in vivo imaging studies is consistent with a reduction in neuropil volume? We develop a framework for addressing these questions and evaluating their implications for understanding the roles of early and late neurodevelopmental influences in the etiology and pathophysiology of schizophrenia.