We tested whether the 2010, 2011 and 2014 European Union bank stress tests produced useful and real information to the market. Using an augmented capital asset pricing model, we analyzed the impact of the information disclosures on each stress test (announcement, methodology and results events) on the stock market returns and risk of banks. Our approach allows an integrated analysis, as a sample of 41 banks that participated in all three stress tests was used. The most significant event was the methodology disclosure, in terms of its impact on risk and returns. In contrast, the results events did not have much impact in the stock market when considering the entire sample of banks. On the other hand, after dividing the sample of banks into two groups (those that passed the 2014 European Union stress test vs. those that failed), we observed a significant reaction of the stock markets in both groups. These findings are consistent with the hypothesis that stress tests provide real and valuable information to the markets about the banking system. A significant part of that information is conveyed by announcement and methodology events.