Learning to read and understand research articles (primary literature) is an important step in the enculturation of higher education students into the scientific community. We presume, based on ideas from the field of genre analysis, that it is important for the development of reading skills to become conscious of the rhetorical structures in research articles. So, we determined how well science students are able to identify 2 important elements of this rhetorical structure: conclusions and grounds. First-year undergraduate life science students who followed a course called "Biomedical Research" made assignments in which they had to identify these 2 elements. We analysed the answers of 20 students in detail and compared their answers with 2 expert readers. Furthermore, we conducted task-based interviews with 4 students to gain more insight into their reading strategies and to determine how they identify conclusions and grounds. Our results show that students and experts defined conclusions and grounds in different ways. Students and experts agreed on the most important conclusion of the articles. However, students identified a wide range of sentences which were not seen as conclusions by the experts. The grounds students mentioned mostly matched their conclusions. Students sometimes failed to mention important grounds for a particular conclusion. In conclusion, our study shows the differences between student and expert readers of primary literature. Based on our results, we formulated criteria for the design of a teaching strategy that aims to improve students' skills for reading primary literature. (Contains 1 figure and 6 tables.)