This report presents the case study of the rapid diffusion of combined heat and power (CHP) units in the Dutch greenhouse horticulture in the period 2003-2009. The aim of the case study is to find explanations for this particular transition, and to generalize on the nature of technology diffusion processes. The study is carried out by means of a literature study, and by means of interviews with sector stakeholders, including several greenhouse growers. In the theoretical framework, technology diffusion is conceptualized as a socio-technical transition, in which the interactions between different actors and the co-evolution of different societal domains form the key characteristics. We adopt a System-Network-Agent approach, as well as the theory of Universal Darwinism, in order to identify and examine the different developments and mechanisms that played a role in the CHP transition. The case study generates three types of insights: observed phenomena, key drivers, and evolutionary mechanisms. These insights contribute to the understanding of the emergence and evolution of technology diffusion in industrial sectors, which may help the formulation of national innovation policy. The five identified key drivers for the CHP diffusion are the opening of the energy market in 2002, the high spark spread during the transition period, the compatibility of output of a CHP unit with greenhouse demand, the flexibility provided by the heat buffer, and the cooperative and competitive greenhouse sector culture. We also found that the CHP diffusion has not been specifically aimed for by the Dutch government, but rather evolved out of interplay between developments in different societal domains. A general conclusion on technology diffusion is that, given the existing variety in social-technical systems and the developments emerging in these systems, each technology diffusion case will necessitate a different degree and nature of government involvement in order for the diffusion to become a success. Therefore, innovation policy makers should consider the co-evolutionary mechanisms inherent in technology diffusion processes.