This paper will outline a pilot study that is part of my ongoing doctoral research into assessment of students’ multimodal text-making in Swedish higher education. In digital environments in higher education students increasingly work with various digital formats, creating multimodal texts that combine modes such as image, moving image, sound and writing. The changing possibilities for students to represent knowledge in multimodal texts also entail changes in how teachers understand and assess knowledge. As a reflection of pedagogy, multimodal assessment practices are therefore closely linked to teaching quality. This presentation aims to describe data, methods and preliminary results from an ongoing pilot study from a Swedish university which offers online and campus-based education. Two teachers, from two courses in different academic subjects and disciplines, have been interviewed focusing on how they recognize and assess four students’ learning and knowledge. Data consists of the students’ digital material and teacher feedback in the Learning Management Platform, and audio recordings of the teacher interviews. Through this fieldwork conducted in a Swedish university, the pilot study offers insights into changing patterns of meaning making and assessment in a digital learning environment, activities which in turn are reflections of teaching qualities. A multimodal social semiotic framework provides tools for studying students’ multimodal texts (written, spoken, visual, gestural). Such multimodal texts are complex representations of knowledge and learning. Improved methods and tools for analyses of these complex multimodal texts would be profitable. To this end, the possibility of designing and developing new analytic software and formats for multimodal transcription is under consideration. In addition to the contribution to understandings of multimodal assessment, this pilot project also therefore seeks to develop multimodal methods for studying complex digital representations of knowledge, to improve teaching quality by facilitating recognition of learning and knowledge in multimodal texts.