Choosing the right time to release a new movie may be the difference between success and failure. Prior research states that the “bigger” a blockbuster is, the more likely it is (and should be) released during a high-demand week. We present a theoretical framework which is consistent with this observation but adds a rather surprising theoretical prediction: among non-blockbuster (i.e., niche) movies, everything else constant, the greater a movie's appeal, the more likely it is released during a low-demand week. In other words, the relation between movie appeal and high-demand-week release is U-shaped: it decreases at low levels of overall appeal (niche movies) and increases at high levels of overall appeal (blockbusters). We provide intuition for this novel result and argue that it is robust to a number of changes in functional form assumptions. We then show that the theoretical results are consistent with the evidence from an extensive data set on international releases. Specifically, we run a series of movie-country-pair regressions with high-demand-week-release as a dependent variable and exogenous shocks to the movie's appeal as an explanatory variable. As predicted by theory, the regression coefficients have opposite signs for the blockbuster and non-blockbuster cases.