Background: One of the key questions of physical education teacher educators (PETE) programmes refers to whether future teachers are prepared to build knowledge and skills to feel self-efficacious in teaching physical education (PE). This issue concerns the instructional model of teaching used to help PE pre-service teachers to master both pedagogical knowledge and motor skills. According to this twofold challenge, the direct instruction (DI) is mainly used for pre-service teacher training. Beyond this traditional model, other instructional models as cooperative learning (CL) approach arise in the initial PE teacher education. Nevertheless, surrounding attempts at innovation, little information related to the instructor's role. Under the social cognitive perspective of self-efficacy and instructional competency building, more information is currently expected with regard to the strategies the instructor uses to scaffold the mastery of skills for PE pre-service teachers' effective teaching. Purpose: The purpose of this article is to consider whether PE pre-service teachers are trained during short training sessions aimed to discover new physical activities. We examine the influence of a scaffolding procedure (CLS design) on PE pre-service teachers' knowledge, skills and self-efficacy in comparison to a CL and a DI experience. This leads to consider to what extent this instructional support provided by the instructor would help pre-service teachers to perceive themselves as self-efficacious to teach contents in PE. Participants and design: After a pre-test, sixty-nine PE pre-service teachers were randomly assigned to one of the three following conditions: CL (14 males and 7 females); CLS (20 males and 8 females) or direct instruction condition (DI; 12 males and 8 females). For the training session a selected CL procedure (Jigsaw) [Aronson, Elliot, and Shelley Patnoe. 1997. "The Jigsaw Classroom: Building Cooperation in the Classroom." 2nd ed. Wokingham: Addison-Wesley Educational]) was used to split CL and CLS participants into mixed-sex teams, whereas DI participants practiced the same exercises in dyads. According to the training conditions, the same instructor provided different information to participants along the three 2-hour instructional sessions with regard to: (a) warm-up (DI), (b) CL organization (CL), and (c) scaffolding integrated into a CL implementation (CLS). Data collection: A Pre-test/post-test design was used to consider PE pre-service teacher's motor skill, knowledge for practice, and self-efficacy improvements. The post-test also examined participants' pedagogical knowledge. Findings: The results showed that the participants in the three conditions progressed on performance, knowledge for practice, knowledge for teaching, and self-efficacy. Although no difference was found in self-efficacy between the three training conditions over time, significant differences appeared on pedagogical knowledge or/and motor skills with an advantage for the CL and CLS participants, respectively. Conclusion: Although short training sessions dedicated to discovering new sports stay problematic for teacher professional development, implementing CL pre-service teacher training designs would be a relevant alternative. Instructional knowledge would be developed mainly when they have explicitly access to information concerning the teacher intervention. Nevertheless, such a scaffolding procedure integrated into CL training designs would need to be applied repeatedly to various physical activities to have an impact on pre-service teachers' self-efficacy.