Purpose: To develop a psychometrically sound screening instrument using a risk and protective factor framework to assess for problematic domains in adolescents at risk of, or beginning to exhibit, substance use behaviors; selected and indicated populations. Four subscales assess the school, family, community, and individual/peer domains. Scores above the domain cut-points indicate the need for further clinical investigation. Methods: The Communities That Care Youth Survey (CTCYS) was used as an item pool, with a convenience sample of 26,781 surveys by students throughout New York State in the year 2000 serving as the sampling frame. Two random twenty five percent samples were drawn to serve as exploratory and confirmatory samples. Principal component factor analysis was run to determine the number of factors that emerged from the data. One item for each factor was included in the screen, and three methods of item selection were compared to determine the approach that best fit this data. Analyses were conducted for concurrent criterion validity, known instrument construct validity, and convergent construct validity. Inter-domain correlations were computed to assess the independence of each domain, and correlations for all gender and race sub-groups in the exploratory sample were examined to gauge the applicability of the scale to these sub-groups. The results were then replicated on the confirmatory sample. Clinical cut-points were determined for ninety five percent sensitivity and tested for positive predictive value, negative predictive value, and specificity. Results: Principle Component Analysis produced thirty two factors and compared favorably to the original factor structure of the CTCYS. The ADSI did a good job of meeting the psychometric standards set. The screen exceeded the concurrent criterion validity criteria set for correlation between the ADSI scale and subscales with its original long form counterparts. Correlations ranged from r = .85 for the school subscale to r = .95 for the overall scale. It produced an adequate correlation with the last 30 day use, the known instrument validity test, of r = .561. The convergent construct validity results, correlation with the Anti-Social Behavior scale, did not meet the criteria set of r = .50, though at r = .494, the results warranted further investigation. The confirmatory analyses supported the exploratory findings, and met the convergent construct validity criteria. The domains of the ADSI showed sufficient independence, indicating that respondents were scoring differentially on domains. The ADSI showed a general applicability across both gender and race and the results were replicated in the confirmatory sample. Cut-points showed sufficient positive predictive values. Implications for practice: The ADSI is an individualized screening instrument that has shown acceptable psychometric properties, is brief, has useful cut-points, and is based on a risk and protective factor model of adolescent substance use. This screen will serve as a first step in a comprehensive assessment of adolescent's who are at risk of, or beginning to show, substance use behaviors.