Epidemiological studies of late-life psychosis pose several methodological challenges and there are few high-quality studies regarding risk factors and prognosis. In the first section we describe some of these challenges: outcome definitions and measurements, overlapping symptoms across various psychiatric outcomes, age of exposure or onset of outcome, timing of exposure in relation to outcome (e.g., mid-life vs late-life exposure), exposure level (”load”), duration and persistence, survival factors, birth cohort effects, study design and analytic strategies, multiple comorbidities, and use of biomarkers. The second section is a review of epidemiological studies of psychosis. Among the key findings: (1) Reports on the prevalence of late-life psychosis are highly variable. (2) Risk factors associated with late-life psychosis include sensory impairment, social isolation, paranoid personality, structural brain abnormalities, cognitive dysfunction, poor physical health, and negative life events. (3) Patients with late-life psychosis have markedly higher mortality rates than the general aging population. (4) New-onset psychotic symptoms in older adults without dementia confers a greater risk for the subsequent development of dementia, but the proportion at risk varies considerably between studies.