This study investigates how six Swedish teachers in upper secondary school apply terms and assessment of written school assignments produced in the course Swedish 1. In addition, the participants’ arguments as to why they are working as they do are highlighted. The study takes on the perspective of phenomenography and an interview study is implemented. The data is analysed according to phenomenography, striving to identify and describe variation. The results show that there are three different ways of working with writing among the participants: 1) texts that are to be graded are only produced in school, 2) texts that are to be graded are mainly produced in school and 3) texts that are to be graded are produced both in and out of school. Four arguments are used as explanations. Reliability and the importance of pupil attendance are used by all participants; equivalence and lack of course time are only used by some. Based on the results, possible advantages and disadvantages of different ways of working with writing are discussed.