Objective: While qualitative research suggests there may be detrimental effects of fitness testing in physical education (PE), little quantitative data have helped to understand how students experience fitness-testing lessons. This study aimed to examine whether the physical self-concept plays a role in how students experience fitness testing at school.Methods: A total of 143 secondary school students (M = 13.48 years) in Western Sydney, Australia, wore an accelerometer during one PE lesson (either a fitness-testing lesson or a 'normal' practice lesson), and self-reported physical self-concept, and positive and negative affect after the lesson. We used linear regression models to determine whether lesson type influenced students' levels of moderate-to-vigorous physical activity (MVPA), and positive and negative affect, and to test whether self-concept moderated the role of lesson type.Results: Lesson type was a significant predictor of MVPA (beta = .11, p = .001), with students being less active in fitness-testing lessons. Lesson type was not a significant predictor of positive (beta = .08, p = .727) or negative (beta = .08, p = .727) affect. However, physical self-concept moderated the effect of lesson type on positive affect (beta = -.42, p = .005), such that those with lower physical self-concept reported lower levels of positive affect when engaging in fitness-testing lessons.Conclusion: While understanding and developing fitness are important elements of PE, group testing of fitness may be associated with poor affective experiences, among those most at risk of being inactive later in life. Therefore, if fitness is to be assessed, new approaches in which students self-assess individual improvement might need to be considered.