DESCRIPTION (provided by applicant): This is the first revision of an application to continue researching the CRA + Vouchers treatment for cocaine dependence. Cocaine dependence is an entrenched U.S. public-health problem. While excellent progress has been made in developing efficacious behavioral therapies for this disorder, there is considerable room for improvement. One promising area for improvement is in matching treatment parameters to patient characteristics so that lower-severity patients not receive more treatment or higher-severity patients less treatment than necessary for a positive treatment outcome. Treatment development for cocaine dependence is often biased towards optimizing efficacy without appropriate attention to constraining costs or reducing costs without sufficient attention to undermining efficacy. Matching the dose of treatment to characteristics that moderate treatment response has the potential to strike a balance between those two biases. We are proposing two matching studies. Study 1 will examine the efficacy of a 6-week contingent voucher intervention with and without voucher values matched to patient severity. We will use marital status (currently married vs. other) and route of cocaine administration (intranasal vs. smoked/intravenous) to classify severity. This severity measure moderates response to the voucher component of the CRA + Vouchers treatment. Cocaine-dependent outpatients will be randomly assigned to CRA plus one of three 6-week voucher conditions: non-contingent vouchers, contingent vouchers set at usual values for all patients, or contingent vouchers set at usual values for low-severity and twice-usual value for high-severity users. We hypothesize a significant interaction between treatment condition and severity, with the efficacy of the 6-week contingent voucher condition being dependent on patient severity when vouchers are set at the usual values, but not when values are matched to patient severity. Study 2 will examine the efficacy of combining 6 weeks of vouchers with CRA delivered as a brief intervention (2 weeks), a moderate-duration intervention (6 weeks), or an intervention wherein duration is matched to two moderators of response to CRA therapy, marital status (currently married vs. other) and alcohol dependence (with vs. without). We hypothesize a significant interaction between treatment condition and severity, with the efficacy of CRA being dependent on marital status and alcohol dependence in the uniform-duration conditions, but not in the condition where CRA duration is matched to patient characteristics. We also are proposing two secondary aims to (a) examine the role played by the behavioral-economic concept of delayed discounting as a potential moderator of treatment response and (b) further understanding of how patient drug-use severity moderates treatment response. Overall, the proposed studies have the potential to advance understanding of how to better match treatment parameters to patient severity in the treatment of cocaine dependence with the goal of streamlining the costs of the interventions without compromising the quality of care offered to more severe patients.