A literature review and power analysis of longitudinal studies published since 1997 investigating the trajectory of neurocognitive deficits across time in patients with schizophrenia was conducted. Ten studies were identified, evaluating a total of 834 patients with mean ages of 24.0-77.8 years at study entry. Power estimates for the 10 studies ranged widely from 0.26 to 0.99 for a medium effect size of 0.4. Despite wide inter-study differences in cognitive measures selected, sample size and phase of illness studied, several consistent themes emerged. Studies of primarily community-dwelling outpatients with schizophrenia revealed that overall measures of IQ and gross cognitive status do not show deterioration greater than that associated with benign aging. Furthermore, performance on specific measures of neurocognition was remarkably consistent across ages studied. Indeed, there is some evidence that IQ, as well as other neurocognitive measures, may show improvement over a 5-year test-retest interval. Findings were consistent whether patients were in their first episode of illness or chronic. These results support a static "encephalopathy" model of cognitive deficits in schizophrenia. In contrast, studies of middle-aged and elderly institutionalized patients with schizophrenia have revealed markedly different findings. There is evidence in this population of a decline in gross measures of cognitive status even over a brief 2.5-year test-retest interval in patients 65 or older that may, in some cases, be linked to the emergence of orofacial dyskinesia. Taken together, these studies suggest two distinct neurocognitive trajectories during the lifespan in patients with schizophrenia that may represent manifestations of distinct pathophysiological mechanisms of the illness during different phases of the disease.