The results of this study are consistent with those of an earlier study that reported shorter lengths of stay among patients with substance abuse disorders than among other psychiatric patients (7). Furthermore, our results suggest the association between substance abuse and shorter stays is maintained even when substance abuse coexists with a serious mental illness.However, compared with psychiatric patients who do not abuse substances, dually diagnosed patients, particularly those who abuse alcohol, are more likely to be rehospitalized after these initial short stays and spend more days in a state facility during a 90-day follow-up period. This increased recidivism was not associated with decreased linkage with aftercare among the substance abusing patients.A primary limitation of this study is the absence of a reliable diagnosis of coexisting serious mental illness. Given the discordance between chart diagnoses of substance abuse and diagnoses derived from the semistructured interview using the CUAD scale, other chart diagnoses must also be viewed with some skepticism. Future studies should include structured interviews to enable independent diagnosis of all psychiatric disorders.Caton and Gralnick (9) have suggested that factors beyond length of stay must be explored to understand the treatment process for psychiatric patients. Looking at the length of only one hospitalization might lead to the inaccurate conclusion that seriously mentally ill patients with substance abuse comorbidities are less expensive to treat. Clearly, a long-term perspective is needed. This study's findings support the need to design and test innovative interventions for mentally ill substance abusers (10).