We examine 269 non-automotive product recall announcements that were published in the Wall Street Journal Index between January 1984 and December 2003. Consistent with previous research, we find statistically significant negative abnormal returns on, and one day prior to, the announcement date. Mean cumulative abnormal returns are not statistically significant over the pre- and post-announcement periods, however, providing evidence in support of the efficient market hypothesis (EMH). These results are robust with respect to the selected index, beta estimation method, and assumption about the behavior of residuals. Moreover, empirical results suggest that important differential industry effects exist and that companies in the drugs/cosmetics industry suffer most from their recall nnouncements.