Urban experiments have been promoted as means to enable innovation for sustainability, particularly in urban mobility. Yet, they have been criticized for struggling to stimulate broader transformations, as they often are detached from public-value principles, lack embeddedness in the cities' everyday realities and are industry-oriented. How cultural changes on different governance levels intersect to produce urban experiments with transformative potential has received little attention. This paper focuses on how urban experiments are co-created with broader governance cultures in multiple governance levels, and what the implications of this co-creation are for urban transformation. We provide a theoretical background on the interrelations between governance cultures and urban experimentation, and the debate on urban experimentation within Science and Technology Studies, transition/innovation studies and urban studies to identify the main barriers for urban transformation. We, then, present our methodology consisting of the case study selection of the multi-level governance nexus State-Region-City in Stuttgart and Karlsruhe, our data collection with interviews and documents, and the analytical tool of storylines to capture the co-production of governance cultures and urban experiments. We continue with the analysis of the case study of automated driving experimentation with the concept of storylines. Our findings show that urban experiments are more likely to lead to urban transformation when the local public sector has a strong role in governance processes, and when experiments emerge through deliberation on daily urban problems and policy agendas. When governance processes are mainly led by state and industry actors who prioritize testing technologies as universal and scalable byproducts, it is less likely for urban experiments to lead to urban transformation. Finally, we discuss when urban experimentation advances technology per se and when it adds public value and advances sustainability, arguing for a co-existence of different kinds of urban experiments. We conclude with future research and policy implications.