Purpose: The purpose of this paper is to present an empirically based discussion of how cooperation between higher education institutions and work organisations (WOs) can increase graduate learning experiences and employability. Design/methodology/approach: Data are based on an electronic and mail-based graduate survey among Norwegian master's degree graduates six months after graduation. 2,232 master degree graduates responded to the survey, giving a 54 per cent response rate. All subject fields were represented. Findings: 77 per cent of the graduates have some forms of interactions with WOs during their studies. The level and mode of participation differ, however, very much, and the benefits of participating differ by mode of interaction. Project-based interaction is positively and significantly associated with completing studies on time. Further, graduates who have participated in either project-based interactions or practice periods have better labour market situation after graduation than their peers. The results hold also when controlling of subject field differences and students' abilities. The study indicates that it is important to look at a wide range of interaction activities and look at the benefits from these activities both in terms of effects on the quality of the learning experience, as well as benefits realised in the transition to work. The data on learning benefits of interaction with WO during higher education are, however, limited, and further studies are needed on this issue in particular. Practical implications: The study indicates that only certain forms of interactions between students and WOs have benefits for students; namely the ones that involve a certain degree of time and commitment. Programme officers in higher education should target more committed interaction forms, because the added value is significantly higher for the students. Originality/value: The study contributes new knowledge about the external networks of higher education institutions, by looking at the role of students in such networks and by measuring the benefits by using a representative sample of students.