In this paper I exploit the removal of Warner Music content from YouTube in January 2009, and its restoration in October 2009, as a plausible natural experiment to investigate the impact of online content availability on album sales. I find that this blackout on YouTube had both statistically and economically significant positive effects on Warner albums, which are quickly moderated as top-selling albums are dropped from the sample. Results also show that albums that have a very successful debut face more displacement from YouTube videos, while the effect on lower debuting albums may be moderated by a promotional effect. (C) 2016 Elsevier B.V. All rights reserved.