Benjamin Speich and colleagues investigate whether rates of trial completion and publication have increased over the past decade, the extent to which non-published trials can be identified in registries, and the association between reporting quality of protocols and premature discontinuation or non-publication of trials.