Purpose: This study aims to focus on living labs as a means of achieving radical innovation by discussing the differences in their network structure and its effect on the type of innovation outcome. Design/methodology/approach: This research analyses 24 living labs in four countries using qualitative methods. Findings: A specific network structure referred to as a distributed multiplex supports radical innovation in living labs, while distributed and centralized network structures support incremental innovations. Also, the results suggest that radical innovation depends on the driving actor and objectives in a living lab. Research limitations/implications: A bias on the perceived novelty of innovation may exist when analyzing data collected through interviews with a limited number of living lab participants compared to a large number of informants. This study proposes a two-dimensional framework based on the network structure to investigate innovation in living labs. Practical implications: This paper offers a classification tool to identify, categorize and make sense of organizations’ participation in open innovation networks and in living labs. Originality/value: The study provides evidence that, although the distributed multiplex network structure supports the emergence of radical innovations, the distributed and centralized network structures support incremental innovation. A combination of a provider- or utilizer-driven living lab and a distributed multiplex network structure, with a clearly defined and future-oriented strategic objective, offers good potential for radical innovation to occur.