Using fixed betting odds for football match outcomes this paper improves on previous versions of outcome uncertainty models of attendance demand by utilizing data on 'repeat' fixtures between two teams. This provides some control over the variation in match characteristics which influence demand. The analysis of 'repeat' fixtures also enables an improved treatment of the 'core support' element of attendances. The contribution of fixed betting odds in explaining attendance demand is confirmed as is the 'U'-shaped relationship between attendance and home team probability of success.