Cognitive-behavioral therapy (CBT) is an effective treatment for alcohol dependence, but little is known about how CBT works to achieve these effects. Although several possible mechanisms have been proposed to explain the effects of CBT, it is rare that more than one mechanism is studied. However, it may be the case that similar outcomes (e.g., abstinence) may be reached through multiple paths. Therefore, essential to conducting work on behavioral change mechanisms is distinguishing different courses or paths and moderating influences (Kazdin, 2007). In the present study, we will focus on 2 key mechanisms posited to underlie the effectiveness of CBT specifically, increasing self-efficacy and self-confidence and reducing positive outcome expectancies for alcohol use, and 2 key mechanisms posited to underlie the effectiveness of a wide range of therapeutic interventions, increasing the therapeutic alliance and reducing/regulating negative emotional states. For the present study, participants will be 72 alcohol dependent men and women who agree to participate in a 12-week trial of CBT for alcohol dependence. In addition, comprehensive research assessments will be conducted with patients at baseline, end of treatment, and 3-months posttreatment.

Specific Aim 1. To examine the within-treatment week-to-week relationship between ratings of 4 key therapeutic mechanisms and alcohol involvement (operationalized as percent days abstinent {PDA} and drinks per drinking day {DDD}) during treatment. It is hypothesized that self-efficacy, outcome expectancies, therapeutic alliance (as rated by the patient and therapist), and negative affect assessed after any given treatment session will each predict alcohol involvement during the following week. Exploratory analyses also will investigate the relationship of a given week's alcohol involvement on each of the four key variables as assessed the following week.

Specific Aim 2. To examine profiles of the four key mechanisms over the course of treatment in relation to alcohol involvement during treatment and during the 3-month follow-up period. These analyses will be descriptive/exploratory. It is expected that profiles will emerge with the key mechanisms that are associated with decreased alcohol involvement (for example, when the alliance is relatively strong throughout treatment or when the alliance grows stronger over the course of treatment, or when positive outcome expectancies grow weaker over the course of treatment, etc.). Similarly, it is hypothesized that profiles will emerge that are associated with little improvement in alcohol involvement (for example, when negative affect is relatively high throughout treatment or when self-efficacy weakens over the course of treatment).

Inclusion Criteria:

• Alcohol Dependent; live within commuting distance of program site; willing to sign informed consent.

Exclusion Criteria:

• Diagnosis of schizophrenia or other psychotic disorder; present with gross neurocognitive impairment; have a current drug use diagnosis other than nicotine or marijuana abuse; have been in substance abuse treatment over the previous 6 months.