Cognitive difficulties are common in people with psychosis and associated with considerable disability. Cognitive remediation (CR) can reduce the burden of cognitive difficulties and improve functioning. While mental health care has predominantly shifted to the community, people with greater illness severity and complexity, and those with poor response to treatment and concomitant greater cognitive difficulties, continue to receive inpatient care. The aim of this study is to review and evaluate the acceptability and efficacy of CR for inpatients with psychosis. A systematic search was used to identify randomized controlled trials of CR for inpatients with psychosis. Demographic and clinical information was extracted by independent raters together with therapy outcomes. Study quality was assessed using the Cochrane Collaboration Risk of Bias Assessment tool. Standardized mean change for cognitive and functional outcomes was calculated using Hedges's g and used to infer therapy effects with meta-analysis. Twenty studies were identified considering 1509 participants. Results from random-effect models suggested that CR was effective in improving processing speed (g = 0.48), memory (g = 0.48) and working memory (g = 0.56). While there was an indication of improvements in the levels of vocational, social and global functioning, these were less reliable. On average, 7% of participants dropped-out of treatment. Studies methodological quality was moderate. CR is an acceptable intervention for inpatients with psychosis and can lead to significant cognitive improvements. Evidence for improvement in functioning requires more robust and converging evidence. Future research should extend the evaluation of inpatient CR to subsequent post-discharge community functioning and further need for care.