The sustainability challenge requires experimentation with innovations, followed by an upscaling process towards a broader regime change in the long term. In Europe we observe various regional hotspots for sustainability experimentation which suggests that there are favorable spatial contexts. Little is known about why different kinds of experiments flourish or fail in various spatial contexts. In this paper we explore these contexts by using the habitat concept. A habitat is regarded as the configuration of favorable local and regional context factors for experimentation. To capture the diversity of these habitats we have constructed archetypical experimentation patterns. These patterns are built up of five dimensions: knowledge, governance, informal institutions, regional innovation advantages, and social learning. In a comparative case study in four city regions in Europe we find a large contrast in habitats. Countercultures play an important role, as they shape a beneficial context for experimentation through alternative ideas and lifestyles. We also find indications that it is important that a combination of several habitat factors is present, and that these factors have aligned and evolved over several years of experimentation, thus leading to a more mature habitat. The research suggests that regional stakeholders can positively influence most of the habitat factors shaping future upscaling. However, there are also some important factors, such as regional knowledge and skills, which have a path-dependent nature and are more difficult to improve in the short term.