School health educators have devoted much attention to cigarette smoking. Recent years have seen the testing of interventions to prevent smoking. To date, controlled studies have not evaluated the added value of skills methods for preventing smoking. This article describes such an evaluation with sixth-grade students from two schools. Subjects were pretested and randomly assigned to receive conventional health education methods or to receive skills intervention. Both conditions included films, peer testimonials, discussions, and homework. Health education condition subjects additionally participated in oral quizzes, games, and debates. Skills condition subjects additionally learned problem-solving, self-instruction, and interpersonal communication methods. At postintervention, skills condition subjects, more than health education condition subjects, had better scores on measures of smoking-related knowledge, attitudes, and intentions. In addition, reported cigarette use, validated by biochemical data collection, was lower in the skills condition than in the health education condition at all postintervention measurements, including a 24-month follow-up. The article discusses the strengths, limits, and implications of the study for other smoking prevention efforts in schools.