Purpose: The purpose of this article is to compare the effectiveness of facilitated networks to other policy instruments for agricultural innovation. Design/ methodology/ approach: In an exploratory study of the Dutch agricultural policy context, we conducted semi-structured interviews with ten experts on networks and innovation. Policy alternatives to networks included research funding, innovation experiments, knowledge vouchers for entrepreneurs, practice networks, competitions for awards/ prizes, innovation subsidies for individual entrepreneurs, legal exceptions, legislation and fiscalisation. Findings: In early phases of the innovation process, facilitated networks were seen as more effective and cost-efficient than the other instruments. This was especially the case for system transformation. However, other instruments can have comparable performance for innovation when they result in sufficient network formation, for instance when they require that target groups build coalitions and other forms of networks. Networks were also seen as effective for system optimisation, but not more cost-efficient than other effective instruments. Practical implications: Past policy experiences with networks enable moving beyond the generic term of "(facilitated) network" to develop more advanced instruments for specific types and phases of innovation. Furthermore, the results suggest that facilitated networks may be a cost-effective alternative to the national extension services of old. Originality/value: Many studies have shown the importance of networks for agricultural innovation. Furthermore, networks offer governments new opportunities to stimulate agricultural innovation. However, less is known about the effectiveness of networks as a policy instrument.