Using the example of the state elections in Saxony in 1999, this article shows how experimental election markets can be used to predict election results.