保護學童避免菸害是公共衛生重要課題，菸害防制教育應盡早施行。本研究依據整合性行為理論設計多元介入方式並評值對國小學童之改善情形。以準實驗研究設計之實驗組對照組前、後測設計，選取臺灣北部偏遠地區八所國小三到六年級學童為研究對象，四所學校為實驗組進行菸害防制介入計畫，另四所學校為對照組持續常規活動。兩組皆於介入前完成前測問卷，介入為期6個月，介入後同步進行後測，於6個月後再追蹤。本研究結果顯示，510位參與學童中，有64.5%學童暴露於二手菸環境；本介入計畫在「菸害知識」、「不吸菸的行為意圖」、「拒吸二手菸自我效能」於控制前測得分的前提下，實驗組顯著較對照組表現良好。本研究支持以整合性行為理論所發展的菸害防制介入計畫具有部分成效，此一整合學生、學校家庭與社區的多元介入方式，有助於作為未來公共衛生推動者菸害防制規劃之參考。 1. Both smoking and exposure to secondhand smoke (SHS) are associated with short-term and long-term health hazards. Protecting school students from smoking and SHS exposure is a key public health topic. Smoking prevention education should be introduced as early as possible to address smoking among students. Studies have suggested that implementing smoking prevention education in elementary schools can be an effective instrument for preventing smoking onset in adolescence. The integrated behavior change model (I-Change Model) is a well-known framework for preventing smoking among youths and adolescents. This model comprises attitude, social influences (comprising norms, modeling, and pressure) and self-efficacy. Although numerous European countries have used this model to implement smoke-free prevention programs, it has not yet been applied in Taiwan. Moreover, smoking prevention programs are lacking at the elementary school level in Taiwan and other countries, especially in rural areas. Therefore, we developed a multicomponent intervention based on the I-Change Model and evaluated its effect on elementary school students in a rural area. 2. In this quasiexperimental study, we applied a repeated-measurement design for the comparison and experimental groups. Students in grades 3 to 6 from four elementary schools in rural Southern Taiwan were enrolled as participants in the experimental group. Students for the comparison group were recruited from four other schools from the same area with similar school type and size as the schools selected for the experiment. In total, the 190 participants in the experimental group underwent a 6-month smoking prevention intervention program, whereas the 330 participants in the comparison group received routine health education. The multicomponent intervention, which was based on the I-Change Model, was focused on smoking and SHS prevention. We applied intervention strategies at the individual, family, school, and community levels. At the individual level, a six-session program (with each session being 45 min) named “Smoke-Free Co.” was implemented. At the school level, the strategies applied included a teacher empowerment workshop and a smoke-free campus program and policy. At the family level, the participants’ parents received a letter on the implementation of a smoke-free environment and smoke prevention worksheets through parent–teacher contact books. At the community level, social media were used to deliver information regarding the implementation of smoke- free environments. Preintervention and postintervention measurements for both groups were conducted through a structured questionnaire. During the 6-month follow-up after the end of the intervention, a third set of responses was obtained from both groups synchronously. Three sets of responses were obtained from 169 and 161 participants from the experimental and comparison groups, respectively. 3. In the present study, almost all the participants (94%) did not smoke, and 64.5% were exposed to SHS. The generalized estimating equation results revealed significant group-related and time-related interactions for knowledge, behavioral intention not to smoke, and SHS resistance self-efficacy. Compared with the comparison group and their baseline scores, the experimental group achieved significantly higher scores for knowledge at the end of the intervention (β = 0.65, p < 0.001) and at the 6-month follow-up after the end of the intervention (β = 0.41, p = .027). For behavioral intention not to smoke, the participants in the experimental group achieved significantly higher scores relative to the participants in the comparison group at the 6-month follow-up after the end of the intervention (β = 0.50, p = 0.045). For SHS refusal self-efficacy, the participants in the experimental group had significantly higher scores relative to the participants in the comparison group at the end of the intervention (β = 4.55, p = 0.010). No significant difference was observed for pros and cons, social norms, social modeling, social pressure, smoking resistance self-efficacy, intention to smoke, SHS exposure, and intention to avoid SHS exposure. 4. In summary, the I-Change Model–based multicomponent intervention (comprising multiple strategy levels) influenced knowledge, behavioral intention not to smoke, and SHS refusal self-efficacy among elementary school students in rural Southern Taiwan. Our results partially correspond to those of other studies that examined the I-Change Model. This slight disparity may be attributed to the cultural differences between different countries and the varying levels of proactivity among teachers. Moreover, during the study period of the present study, some participants graduated from elementary school and progressed to junior high school, and this transition could have reduced the effects of the intervention. Therefore, the program must incorporate other effective strategies to enhance its effectiveness. This study has several limitations. First, selection bias may have influenced the results; however, generalized estimating equations were used after controlling for differences in demographic variables among the two groups to address the aforementioned concern. Second, the lower retention rate of our comparison group could also have affected the results. Third, the collected self-reported data could have caused measurement errors because of social expectations. To eliminate the effects of this bias, we collected anonymous data and instructed the participants to complete the questionnaire on the basis of their actual situation. 5. The current findings suggest that, for children who do not smoke but are exposed to SHS, the I-Change Model–based smoking prevention intervention was partially effective in influencing knowledge, behavioral intention not to smoke, and SHS refusal self-efficacy. This program was appropriately applied to students in grades 3 to 6 in rural Southern Taiwan. Our results relating to the integration of multiple intervention methods (for students, schools, families, and communities) can serve as a reference for public health practitioners with respect to prevention planning. In the future, the effectiveness of interventions involving smoking prevention programs must be enhanced by increasing their implementation intensity and utilizing social media. Furthermore, these measures can act as a booster program with a protective factor that helps student to cope with the pressure of school life and confront challenges when they progress to junior high school.