This study investigates the effectiveness of using peer review in the context of scripted collaboration to foster both domain-specific and domain-general knowledge acquisition in the computer science domain. Using a one-factor design with a script and a control condition, students worked in small groups on a series of computer science problems requiring both domain-specific and domain-general knowledge. The scripted groups carried out peer reviews of their draft solutions. Scripted group members spent more time developing their solutions and reported a higher subjective learning outcome than control group members. Scripted groups acquired also significant domain-general knowledge on peer reviewing; however, there were no substantial differences with respect to the acquisition of domain-specific knowledge (as compared to control groups). Overall, the study provides evidence that students in computer science can benefit from peer review-based scripted collaboration and acquire domain-general knowledge without any harm for domain-specific knowledge. The issue of why domain-specific knowledge was not also improved is discussed from various viewpoints. (Contains 1 note, 3 figures and 7 tables.)