The paper brings Friedman's (1971) collusive game to data and investigates whether the merger between the fifth and fourth largest brewer (G. Heileman and Stroh) of the US beer industry in the mid 1990's had a significant impact on the incentives to collude in the industry. It does so by firstly estimating a random coefficient Logit demand system for the US beer market. In a second step the demand estimates are used to conduct a merger simulation (Davis, 2006) quantifying coordinated effects of the merger. The results show that the change in the likelihood of collusion for the non merging parties was negligible, but significantly increased for the merged party.