Over one hundred thousand Veterans annually utilize mental health residential rehabilitation programs (RRTPs) and psychosocial rehabilitation and recovery centers (PRRCs). Through support and comprehensive programming, these mental health rehabilitation milieus help Veterans with mental illness overcome barriers to community reintegration. Despite the success of these milieus, cognitive impairment is commonly observed in Veterans with mental illness, reduces gains from programming, and limits successful recovery. Cognitive impairment is a common transdiagnostic illness dimension conferred both directly and indirectly by mental illnesses. Cognitive rehabilitation through remediation strategies can potentially attenuate cognitive impairment for these Veterans and improve outcomes, but there are two inter-related problems that limit the effectiveness of such interventions. First, we are not able to identify which Veterans will benefit from mental health rehabilitation programming at program entry. Second, we are not able to predict which Veterans are going to benefit from any specific cognitive remediation intervention. This CDA-2 application seeks to test whether an electroencephalographic (EEG) biomarker, mismatch negativity (MMN), can be used to predict Veteran recovery in mental health rehabilitation treatment settings, and identify Veterans who will respond to cognitive remediation interventions. MMN is an event-related potential which is considered to be a biomarker of information processing, linked to cognition in healthy subjects and cognitive impairment in a variety of neuropsychiatric illnesses. MMN also mediates psychosocial and functional outcomes in individuals without any psychiatric comorbidities and individuals with mental illness. Recent work also suggests that MMN can identify individuals who will experience gains from a full course of cognitive remediation when measured over the first hour of cognitive remediation. These biomarker relationships (cognition, cognitive remediation sensitivity, functioning) have not yet been definitively established in a heterogenous Veteran population receiving care in real-world settings like RRTPs and PRRCs. Veterans with mental illness will be recruited from the VA San Diego Healthcare system RRTP and PRRC at program entry. Baseline measures of functioning, psychosocial disability, cognition and treatment engagement will be collected. Following these assessments, Veterans will undergo testing to collect MMN data, and then will be challenged with a one-hour cognitive remediation exercise, which is a typical component of full multi- hour cognitive remediation programs. Veterans will be followed with monthly assessments of psychosocial disability and treatment engagement. At the end of study, functioning will also be re-assessed. If successful, results from the studies proposed will create an objective, precision-medicine platform which could fundamentally change how RRTP and PRRC approach rehabilita...