Background Children are increasingly exposed to stressors related to urbanization, climate instability, and biodiversity loss, which may contribute to poor well-being. School-based outdoor education has shown potential to support both well-being and learning, but robust evidence from randomized trials remains scarce. Science education outdoors, by engaging students with local biodiversity, may enhance scientific observation skills while contributing to public health goals such as stress reduction and connection to nature. Methods This cluster randomized controlled trial (cRCT) will compare outdoor versus classroom-based science instruction among 5th- and 6th-grade students in primary schools in the greater Montreal area (Québec, Canada). Forty classes (≈1,000 students) will be recruited across two waves (2025–2026). Teachers will be randomized (1:1) to teach life science content either outdoors (intervention) or indoors (comparator) for 2 h per week over 5 weeks. The primary outcome is students’ scientific observation skills, assessed through structured interviews and a validated observation framework. Secondary outcomes include eco-anxiety, perceived stress, connection to nature, and basic psychological needs satisfaction measured with standardized questionnaires. Analyses will use mixed-effects regression models to account for clustering at the class and school levels. Discussion This study will implement a cluster randomized controlled design to examine the effects of outdoor education on students’ science learning, stress, eco-anxiety, and psychosocial factors. By embedding the intervention in the regular science curriculum, the trial aims to generate ecologically valid evidence relevant to both education and public health sectors. Despite the relatively short intervention duration, the study is designed to inform intersectoral initiatives addressing health-promoting school environments and to generate evidence regarding the role of nature-based learning in urban settings. Clinical trial registration ClinicalTrials.gov , identifier NCT07147634.
Ayotte-Beaudet et al. (Thu,) studied this question.