Previous analyses of dairy farm structural change focus on the variation over time in one or a small number of regions. We present an EU-15 cross-regional analysis of the development of dairy farm numbers in different size classes over the period 1995–2005. Our purpose is to measure the explanatory relevance and effect of key factors suggested in the theoretical and empirical literature on structural change. Apart from the unprecedented scope, the underlying Markov chain analysis also contributes by combining observed transitions in micro (farm level) data with macro (sector level) data on farm numbers. Results show widely significant impacts of most considered explanatory variables, but also reflect and illuminate the complexity of the underlying processes.