Contribution: This article provides evidence on the effectiveness of game-based learning (GBL) for computer science education when using educational video games created by teachers using authoring tools. Background: Although teacher-oriented authoring tools for creating educational video games can help overcome the main barriers hampering the use and uptake of GBL, there is a lack of studies examining the effectiveness of educational video games authored by teachers using these authoring tools by means of rigorous scientific methods. Research Questions: Is GBL using teacher-authored games more effective than traditional teaching in terms of motivation for computer science students? Is GBL using teacher-authored games more effective than traditional teaching in terms of knowledge acquisition for computer science students? Methodology: Two randomized control trials with pretest, post-test, and one questionnaire were conducted to compare the effectiveness of traditional teaching and GBL in two computer science courses. The sample of the two experiments was composed by 75 and 49 students, respectively. Half of the students attended a traditional lecture, while the other half learned solely by playing teacher-authored educational video games. Findings: The results show that GBL using teacher-authored games was practically as effective as traditional teaching in terms of knowledge acquisition, but that it was emphatically successful in increasing student motivation. Students who learned by playing educational video games found the experience more motivating and fun than their counterparts, and a vast majority of them preferred the GBL approach over traditional teaching.