This paper investigates whether there is any consistency between banks' financial strength ratings (bank rating) and their risk-return profiles. It is expected that banks with high ratings tend to earn high expected returns for the risks they assume and thereby have a low probability of experiencing financial distress. Bank ratings, a measure of a bank's intrinsic safety and soundness, should therefore be able to capture the bank's ability to manage financial distress while achieving risk-return efficiency. We first estimate the expected returns, risks, and financial distress risk proxy (the inverse z-score), then apply the stochastic frontier analysis (SFA) to obtain the risk-return efficiency score for each bank, and finally conduct ordered logit regressions of bank ratings on estimated risks, risk-return efficiency, and the inverse z-score by controlling for other variables related to each bank's operating environment. We find that banks with a higher efficiency score on average tend to obtain favorable ratings. It appears that rating agencies generally encourage banks to trade expected returns for reduced risks, suggesting that these ratings are generally consistent with banks' risk-return profiles.