This research investigates how methods of learning influenced the emerging wind power industries in the Netherlands and Denmark. It is found that the manufacturing and implementation successes in Denmark contrast with the relatively poor progress in the Netherlands, and that one of the reasons for this is the contrast in learning mechanisms between the Countries. We start from the perspective of innovation systems. Within these systems we place the focus on four types of learning processes: learning by seat-china. learning by doing, learning by using and learning by interacting. It is concluded that in Denmark, learning by interacting was the most important learning process, while in the Netherlands it was learning by searching. The Dutch wind turbine innovation system was a typical 'science-push' innovation system. The aim was to develop large wind turbines at a fast pace, based on the results of scientific research. Because of the lack of contacts between the researchers and the wind turbine producers, the implementation of the research results was problematic. Contrarily, in Denmark the focus Was on knowledge transfer between turbine producers, turbine owners and researchers. In this innovation system, conditions for learning by interacting were optimal. In this way, wind turbines were successfully, though slowly, scaled up and improved. (C) 2003 Elsevier Ltd. All rights reserved.