Dual diagnosis patients come to treatment with a variety of deficits,talents, and motivations. A biopsychosocial treatment plan involves multiple interventions, including medications, medical treatment, psychotherapy, family therapy, housing, and vocational rehabilitation. Treatment must be individualized and integrated, and this requires collaboration among a variety of health caregivers. There is empirical evidence that dual-diagnosis patients can be helped to stabilize, to remain in the community,and even to enter the workforce. Behavioral interventions are key ingredients to integrated and comprehensive treatment planning. There is no single model for dual disorders that explains why substance use and psychiatric illness co-occur so frequently. Mueser et al described four theoretical models accounting for the increased rates of comorbidity between psychiatric disorders and substance use disorders. They suggested that there could be a common factor that accounts for both, primary psychiatric disorder causing secondary substance abuse, primary substance abuse causing secondary psychiatric disorder, or a bidirectional problem, where each contributes to the other. There is evidence for each, although some are more compelling than others, and none is so compelling that it stands alone. Although family studies and genetic research could explain the common factor, no common gene has appeared. Antisocial personality disorder has been associated with very high rates of substance use disorders and mental illness; however, its prevalence is too low to explain most of the co-occurring phenomena. Common neurobiology, specifically the dopamine-releasing neurons in the mesolimbic system, also may be involved in mental illness, but this is not compelling at the moment. The Self-medication model is very appealing to mental health professionals, as an explanation for the secondary substance abuse model. Mueser et al suggest that three lines of evidence would be present to support this explanation: (1) patients would report beneficial effects of substance use on their symptoms; (2) epidemiology would report that a specific substance would be used by specific psychiatric disorders, and (3) psychiatric patients with severe symptoms would be more likely to abuse substances than those with mild symptoms. Unfortunately the research data do not support these. The primary substance abuse causing secondary psychiatric disorder model could be explained by neuronal kindling from substance-induced disorders. Patients who develop the psychiatric disorder after the substance use disorder do have a course of illness similar to those with a psychiatric disorder, but without substance use disorder. The bidirectional model is consistent with the tendency of disturbed teenagers to socialize with youth using alcohol and drugs; however, this model has not been tested rigorously in research studies. With such a disparate set of models, behavior interventions are conceptualized best as a multi-component program, a treatment plan that generates a problem list and devises an intervention to respond to each member of the list. This requires a talented, multi-disciplinary team or network that can assess carefully and package the interventions creatively, and dose the treatment components empathically to fit the patient's tolerance, motivation, and abilities.