This paper examines four learning processes that occur simultaneously during wind turbine development and diffusion: learning by searching, learning by doing, learning by using, and learning by interacting. It considers how these different learning processes affect technology development, which factors facilitate or impede these kinds of learning processes, and how they can be facilitated better to accelerate technological change. These questions are addressed to a historical case study of large-scale and small-scale onshore wind power development in The Netherlands, with some provisional conclusions drawn from recent research on the development of offshore wind. The most prominent learning process in onshore wind power development in The Netherlands was learning by searching. However, no reliable wind turbines, let alone a wind turbine industry that was able to compete at international level, were produced as a result of this learning process. Other learning processes were needed, like learning by doing and learning by interacting with the other actors involved. In the recently set up offshore wind-power innovation system, we see some learning from past mistakes. Wind turbine users are participating more actively and prominently than in the onshore innovation system, allowing more learning by using and learning by interacting. Furthermore, the focus is less on rapid up-scaling and high-tech innovation, but more on learning with existing technology in a different, offshore, environment.