Because the importance of coalition formation has been long established, there is no shortage of ideas explaining and predicting coalition outcomes. However, one of the problems for contemporary coalition research is that most coalition theories have been thoroughly tested on the same data on national governments that have formed in the West European postwar democracies. This stresses the need for finding new data that can increase our ability to test and refine coalition theories. This article uses unique elite survey data from an investigation conducted among councilors in a large sample of local authorities in Sweden to test hypotheses on coalition formation. Another problem in coalition research is that the large number of coalition hypotheses have not been tested extensively using multivariate models that provide sufficient controls. By using a methodological approach that models government formation as a discrete choice between potential governments, we can draw conclusions about the relative importance of different types of variables. The results found in this analysis indicate that we have to pay attention to both traditional variables, such as size and policy, and institutional variables if our aim is to explain and predict coalition formation.