The "expert blind spot" (EBS) hypothesis implies that even some experts with a high content knowledge might have problems in science communication because they are using the structure of the content rather than their addressee's prerequisites as an orientation. But is that also true for students? Explaining science to peers is a crucial part of cooperative learning methods such as the "jigsaw method". Our study examined the relationship between science communication competence (SCC) and content knowledge (CK) of 10th-grade students (N = 213). Using latent class analysis, we identified two types of students with a different relationship between CK and SCC. Using path analysis, we found that the first type of 109 students primarily used their science CK as the "resource" for addressee-oriented science communication and both their SCC and their CK were correlated with each other. For the second type of 104 students (who used other resources), their CK even had a small negative effect on their SCC. Using t tests, we found that those students primarily using their CK as the resource for communication performed significantly worse in the communication test than did those students who used other resources. Using the EBS hypothesis, we suggest that students' CK might have ambiguous effects in communication if the content structure--rather than their addressee's prior knowledge--is used as the primary orientation for communication. We suggest that an effective use of cooperative learning techniques in classroom requires a special prior training for their science communication skills.