This study examines the design and use of task-specific rubrics in a Swedish secondary school setting, against the backdrop of the theory of formative assessment and feedback. For teachers, rubrics can be a tool for systematic analysis of curricula and grading criteria (so called knowledge requirements), while for students, rubrics can highlight the connections between learning objectives, teaching and assessment. The purpose was to study the design of rubrics as well as teachers’ beliefs and opinions about and experiences of using rubrics. A phenomenographic approach was used, supported by Borg’s theoretical framework on teacher cognition and combined with qualitative content analysis. The empirical material for this study consisted of seven examples of rubrics, and transcripts of semi-structured interviews with five teachers of Swedish. The analysis of the rubrics showed that, from a usage perspective, these can be divided into three groups: rubrics for summative assessment, rubrics for formative assessment and rubrics that appear suitable for both formative and summative assessment purposes (regardless of actual use). The rubrics studied could also be divided into two main groups from a design perspective: teacher-made rubrics not clearly modelled on official texts, which were rare, and rubrics based either on the grading criteria for the relevant course or on task-specific rubrics derived from national tests, both of which were equally common. Furthermore, three relatively clear perspectives were found in the transcripts of the teacher interviews: the teacher’s, the student’s and society’s. These perspectives were shown to be interdependent. The interviews illustrated that teachers use rubrics to clarify the learning objectives, to improve students’ learning or performance, and/or to help students understand teacher feedback. From the teachers’ own perspective, possible motivations for using rubrics were 1) helping the students understand the learning objectives, 2) planning their teaching and assessment activities, and 3) giving feedback. Teachers did, however, hold partly differing views on which of the potential ways of using rubrics were best suited to their own teaching and to their students’ learning.