The purpose of this study was to examine a hybrid combination of sport education and the step-game-approach (SGA) on students' gameplay performance in volleyball, taking into account their sex and skill-level. Seventeen seventh-grade students (seven girls, 10 boys, average age 11.8) participated in a 25-lesson volleyball season, in which the structural features followed those of sport education and the specific content followed the didactical framework of the SGA. Data were collected prior to the first lesson (pre-test), following the completion of the unit (post-test) and seven days after the post-test (retention). The game performance assessment instrument was used in a systematic observation of video records of students' behaviours while playing a 10-minute 2vs2 game. A 2 (sex) × 3 (time) and a 2 (skill-level) × 3 (time) analysis of variance (ANOVA) with repeated measures across the three tests was used to compare the students' performances on all indexes. Results showed that both boys and girls made improvements from the pre-test to the post-test and maintained those improvements to the retention test in almost all the indexes analysed. Further, the results also showed that lower skill-level students realized greater gains than those of higher skill during the unit, which reinforces the idea that future implementations of this hybrid approach should adjust content and learning task to different skill-levels.