This study assesses the implementation quality of Think Smart, a school-based drug prevention curriculum that was designed to reduce use of harmful legal products (HLPs; e.g., inhalants and over-the-counter drugs), alcohol, tobacco, and other drugs among 5th- and 6th-grade students in frontier Alaska. Participating in the study were eight communities that took part in a larger randomized control trial to assess the short-term effects of the Think Smart curriculum. Video-recorded observations of the 12 core and 3 booster lessons were conducted in 20 classrooms. Ninety-five sessions were randomly selected from 228 usable videodiscs, and two pairs of researchers observed each video recording to code level of dosage, adherence to curriculum design, and teachers' delivery skills. Inter-rater reliability for all implementation quality measures was very high. An expert panel consisting of 16 scientists reviewed the results of the implementation study and concluded that the level of dosage and adherence to the curriculum design was at least as high as those yielded by similar studies. However, the panel assessed the delivery quality to be only marginal in comparison to results of other studies. The experts concluded that the implementation quality of the Think Smart curriculum was adequate even though the teachers' delivery skills were only marginal. A bootstrapping analysis, in which 1,000 samples were drawn for each implementation quality result, found the expert judgments to be reliable. The authors conclude that despite some limitations, video-recorded observations, as well as expert judgment, provide strong methodologies that should be considered for future implementation quality studies.