A statewide campaign to reduce youth tobacco use has resulted in 78 of North Carolina's 115 school districts adopting a 100% Tobacco-free school policy (TFS). Research was conducted to determine whether three dimensions of policy implementation - policy communication, compliance monitoring, and policy enforcement - were associated with student smoking, and to elucidate the policy implementation process in middle and high schools across the state. Sixty-five principals from middle and high schools that had a 100% TFS policy and that had participated in the 2005 Youth Tobacco Survey (YTS) completed a survey of their schools' implementation of the 100% TFS policy. Using data from the survey, overall implementation ratings (IRs) that demonstrated the level of policy implementation of the 100% TFS policy at each school were created. The IRs were based on the schools' achievement of a set of eight objectives related to the three dimensions of policy implementation. The average school IR was 7.05 out of 9.0, with a range of 4.8 - 9.0. No correlation between IRs and current student smoking was found. Multiple regression was used to explore whether the communication, compliance monitoring, and enforcement subscale scores of the IR predicted the current student smoking rate at the school. A moderate positive association was found between the current student smoking rate and the communication subscale of the implementation rating. Further analysis was conducted on the various subscales. A one-way analysis of variance found a positive relationship between the amount of time that that a school has had a policy in place and the level of compliance monitoring at the school. Three categories of policy compliance violations were also analyzed: (a) the number of policy violations reported by the principal for the academic year, (b) student reports of smoking on campus or at campus events in the 30 days prior to taking the survey, and (c) student reports of school staff smoking on campus or at school events in the thirty days prior to taking the survey. A small negative correlation was found between the IR and the total number of policy violations for the academic year (as reported by principals) at both high schools and middle schools and a moderate negative correlation was found between the enforcement subscale and the number of policy violations at middle schools. A moderate negative correlation was also found between the percentage of students who reported smoking on ...