We estimate the effect of the spring 2023 failures of Silicon Valley Bank and Signature Bank on the “connectedness” of US bank stock return volatilities using the forecast error variance decomposition framework of Diebold and Yilmaz (2012, 2014) and Lastrapes and Wiesen (2021). Using split-sample and time-varying VAR methods, we find that those failures significantly increased spillovers across a sample of surviving regional banks, but had only small and temporary effects on spillovers across systemically important too-big-to-fail banks. Our main findings imply that regulatory policy toward systemically important banks has been credible but that additional oversight of regional banks should be considered.