Team Teaching Models in Primary Physical Education: Effects on Basic Motor Competencies and Self-Reported Physical Literacy

<b>Background/Objectives:</b> To address the inconsistent provision of specialist physical education (PE) in primary schools, this study investigated the comparative efficacy of distinct team teaching configurations. The objective was to compare these instructional models' impact on students' basic motor competencies (MC) and self-reported physical literacy (PL). <b>Methods:</b> This cluster-randomized trial involved <i>N</i> = 266 students (grades 1-4) in Slovakia, assigned to one of five instructional models (PE teacher; GT+PET; PET+AT; GT+C; and GT+AT). The five-month intervention included two 45 min PE lessons weekly. Given the cluster design and non-normal data, the Wilcoxon singed-rank test and Kruskal-Wallis H-test were applied to assess the differences, and Cohen's r was applied to determine effect size. <b>Results:</b> Comparative analysis showed no significant differences across teaching models for Self-Movement (<i>p</i> = 0.544), Object-Movement (<i>p</i> = 0.138), or PL (<i>p</i> = 0.219). Significant within-group MC gains were found in 4 teaching models, yet the practical effect size was generally weak to moderate (r ranging from 0.21 to 0.69). The strongest practical improvement was observed in the AT+PET Self-Movement group (r = 0.69). In contrast, the GT+PET configuration achieved no significant MC gain. For PL, only the AT+PET and GT+PET models showed significantly positive but moderate changes (r = 0.32 and 0.37). Conversely, the GT+C model resulted in a moderately significant decline in PL (Δ = -9.16, r = 0.43). A positive but practically weak correlation emerged between the MC subscales and PL (ρ ranging from 0.135 to 0.238, <i>p</i> < 0.05), with the highest limited association for Catching (ρ = 0.377, <i>p</i> < 0.01). The frequency of organized out-of-school physical activity was positively correlated with MC subscales (ρ = 0.195-0.282, <i>p</i> < 0.01) but not significantly correlated with PL. <b>Conclusions:</b> No single teaching model proved superior for improving overall motor competence or self-perceived physical literacy. While most effective configurations yielded moderate practical gains, the GT+C model presents a key paradox: while effective for objective skills, it proved detrimental to self-perceived physical literacy. These findings lead to explicit policy and implementation recommendations focused on strengthening collaborative instruction. Policymakers should consider strategies to support the integration of specialist PE teachers (PETs), such as establishing co-teaching as a recommended practice and allocating dedicated resources for funding and collaborative planning time to leverage the specialized knowledge they bring. Furthermore, schools are encouraged to focus on the effective implementation of PET-involved team teaching approaches (e.g., AT+PET and GT+PET). These models are not only effective but also support the co-professionalization of the generalist teacher, which is essential for ensuring that high-quality, evidence-based PE practices are consistently embedded.