In this article, we examine a database assembled from an Australian public register of 553 merger decisions taken between March 2004 and July 2008. Mergers may be accepted without public assessment, accepted in conjunction with publication of a Public Competition Assessment, or rejected. The public register contains qualitative information about the reasons given by the regulator for each decision. We estimate an ordered probit model, using these three possible outcomes, with the objective of gaining a better insight into the regulator's decision-making process. Our two major findings are: (i) the existence of entry barriers and the existence of undertakings are highly correlated with the regulator's decision to closely scrutinise a merger proposal; and (ii) if we compare two decisions, one which does not mention entry barriers (or import competition) with a decision that does mention entry barriers (or import competition), then the latter is significantly more likely to be opposed than the former.