Background and Context: Pair programming is an important approach to fostering students' programming and collaborative learning skills. However, the empirical findings on pair programming are mixed, especially concerning effective instructional design. Objective: The objective of this literature review is to provide lecturers with systematic knowledge of current research evidences on the implementation of pair programming in courses by outlining the current state of knowledge on the effects of pair programming in higher education and on effective instructional design. Method: The results are based on an analysis of 61 articles conducting empirical studies on pair programming in higher education published between 2010 and 2020. Findings: Results of studies on the effects of pair programming in comparison with solo programming are to a great extent positive. Regarding instructional design there remain open questions, such as on effective instructional guidance and appropriate task design. Implications: We highlight the need for more research on instructional design of pair programming to provide, for example, knowledge about effective procedures for pair programming, effective guidelines, or problem-solving approaches when problems occur within a team. We give recommendations for practitioners based on our findings.