The pricing of Initial Public Offerings (IPOs) in the short-run has been analyzed by several theoretical and empirical studies referring to the major international stock markets; recently this topic has been particularly popular in Europe due to the high-tech and Internet stock IPOs euphoria (partially flopped in 2000) and to the increasing number of firms going public as well. This paper presents an empirical study conducted on a unique survey of 164 IPOs on the Milan Stock Exchange between January 1985 and August 2000. In particular we aim at determining the driving forces of IPOs initial and short-run market performance. First, we analyze the first-day abnormal return. We find a significantly positive underpricing, equal to 23.94%; nonetheless, 25% of IPOs in our sample are initially overpriced. Second, we separately consider fixed-price IPOs (basically between 1985 and 1994) and IPOs with book building (mostly between 1995 and 1999): we find significantly different levels of underpricing (28.33% vs. 8.12%), and an informative role of revisions in the filed price range, this supporting the “information gathering” and “partial adjustment” theories. Thus we stress the importance of choosing adequate IPO placing strategies in order to reduce information asymmetries between the market and the investors. Third, we try to point out proxies of information asymmetries influencing the initial underpricing. For fixed-price IPOs we find a negative correlation between the underpricing and the age and systematic risk of the firm, and a positive correlation between the underpricing and the market index momentum and volatility. Coherently with the findings above, with book building the correlation with the market sentiment still remains, but the age of the firm is no more significantly correlated; all information collected by the offering parties are endogenized in the revision of the file price range, which is informative for investors. Therefore, we contend that book building allows investors to gather information at a lower cost: they obtain public information by looking at the market momentum and may extract private information from the revision of the prospectus price range as to require a lower underpricing. Finally, we look at the IPOs performance and trading volume in the first weeks of listing. We find that the initial returns contain almost all the underpricing. While the buy-and-hold performance of hot IPOs does not sensibly change after the listing, we highlight that cold IPOs move to negative returns, this suggesting temporary price support activity by underwriters. We verify that underwriters manage over-allotment and green shoe options and buy IPOs shares in the after-market in order to avoid negative initial returns. J.E.L Classification codes: G30, G32.