Peer review is a cornerstone of scientific publication. However, it is time consuming for reviewers and contributors, and must be balanced with editorial oversight for balance and bias. To test a more efficient method of reviewing manuscripts, we performed a randomized trial comparing traditional peer review for all manuscripts received by the Annals of Neurology with an early screening approach in which six editors rejected a manuscript without external review when the chance of acceptance was deemed very low. Of the 351 manuscripts entered into the trial, 88 were randomized to traditional external review and 263 to early screening. Rates of final acceptance were similar in the two groups (p = 0.41). Final decisions were more delayed for traditional review (mean 48 days versus 18 days with early screening; p < 0.0001) and more reviewers were required for each manuscript (mean 2.3 versus 0.7 with early screening; p < 0.0001). Among accepted manuscripts, reviewer ratings of scientific and clinical impact were similar. We conclude that a method of early screening of manuscripts for appropriateness for publication results in substantial decreases in the time between manuscript submission and publication decisions, and reduces the burden on reviewers with minimal impact on the quality of accepted manuscripts. Editorial screening is now journal policy.