Pilot- and demonstration plants (PDPs) represent bridges between generating basic knowledge and technological breakthroughs on the one hand, and industrial applications and commercial adoption on the other. This paper reports on a longitudinal study of how two technological fields that received significant public funding evolved—biochemical conversion of biomass and thermal conversion of black liquor. In doing so, this study makes two contributions. First, it provides a framework for analyzing the roles of various types of PDPs in developing new technology. The framework highlights the learning processes taking place at and around these plants and how they contribute to reducing different types of risks. It also elaborates on the importance of actor networks and institutional preconditions, and how both network performance and institutions can be influenced through various strategies. Second, the article contributes with new insights into the challenges of innovation policy in a PDP context. A policy mix is often required because policy cannot be considered meaningfully at a single level of government and will therefore be influenced heavily by limited foresight and politics (both nationally and locally). Therefore, policy must address both the need for parallel and iterative public funding of R&D and different types of plants, as well as attempts to directly influence collaborative processes in actor networks.