Schizophrenia, schizoaffective disorder and bipolar disorder are serious, chronic mental illnesses whose prevalence, severity, and costs render them major public health problems in the United States and throughout the world. Research has demonstrated that approximately 50% of people diagnosed with a serious mental illness (SMI) also carry a diagnosis of substance abuse or dependence (dual diagnosis). Dual diagnosis is associated with more negative outcomes compared to persons diagnosed with serious mental illness alone. Prior research indicates that token economy programs, which are social learning programs that utilize operant conditioning principles to affect individual behavior, may enhance patient functioning, treatment adherence, and outcomes during psychiatric inpatient hospitalization, in addition to helping severely mentally ill patients function more productively in society after discharge. This study sought to investigate differences between two groups of inpatient veterans- inpatients with serious mental illness and dually diagnosed inpatients- with regard to performance in and response to the Points Incentive Program (PIP), a token economy program implemented by the VA Maryland Healthcare System. The study was a retrospective analysis of archival data obtained from male veterans hospitalized on an inpatient unit from 2002-2006 (N = 147). The results showed that dually diagnosed veterans functioned as well as SMI veterans in the PIP. Results also indicated that PIP points per day appeared to be a predictor of length of time until re-hospitalization, even after controlling for diagnostic group. Therefore, a conclusion may be made that token economy programs such as the PIP that utilize social learning principles can potentially influence patient behavior and functioning in society after discharge.