Purpose – The purpose of this paper is to re-examine the issue of financial analysts' reaction to enterprise resource planning (ERP) announcements by employing actual firm data and archival earnings forecast observations. As an extension of prior ERP studies, this paper also tests whether forecast revisions vary with the timing of adoption. Design/methodology/approach – Based on 188 firms that announced ERP plans during the years 1993 through 2002, this paper investigates the financial analysts' reaction to ERP announcements by comparing their earnings forecasts issued immediately before and after the ERP announcement. To examine the effect of adoption timing, this paper partitions the sample into three groups: early (1993-1997), middle (1998-1999) and late (2000-2002) adopters. Findings – Results show that significantly positive revisions occur in longer term forecasts (i.e. three-year-ahead forecasts) but not in the shorter term predictions such as one- and two-year-ahead forecasts. In addition, there is some weak evidence that financial analysts react less positively to middle adopters than to early or late adopters. This finding could be attributed to the fact that many firms adopted ERP systems to work out the Y2K problems during the 1998-1999 period. Originality/value – The main finding confirms that financial analysts consider ERP implementations beneficial to the adopters in the long term. Companies contemplating ERP adoption should take this time horizon into account.