The growing importance of medical imaging in everyday diagnostic practices poses challenges for medical education. While the emergence of novel imaging technologies offers new opportunities, many pedagogical questions remain. In the present study, we explore the use of a new tool, a virtual microscope, for the instruction and the collaborative learning of pathology. Fifteen pairs of medical students were asked to solve diagnostic tasks in a virtual microscopy learning environment. The students' collaborative efforts were analysed on the basis of approximately 20 hours of video recordings. Our analyses show how students use the technology as a mediating tool to organize, manipulate and construct a shared visual field, and later, shared understanding of the problem and solutions. Organization of the visual field is done through multimodal referential practices: gestures, three dimensional manipulation of the image and paced inspection of the specimen. Furthermore, we analyse and describe how the aforementioned practices coincide with students' medical reasoning in this particular learning context. The analysis of medical students' diagnostic work illustrates the collaborative potential of the virtual microscopy environment and how such interactive tools render the traditional distinction between collaborating around or through computers irrelevant, as even face to face collaboration becomes enacted through technology. Finally, we argue that as technologies develop, understanding the technical side of image production, or any representation, becomes an integral part of the interpretative process. How this knowledge is communicated to the students may play a substantive role in how students learn to interpret medical images.