Abstracts (and other short surrogates for the complete text) of a document have been generally considered a valuable aid to the reader in quickly determining the relevance of a document (as well as, in some cases, serving as a substitute for reading the text, and as a separate surrogate in secondary services). A test of this hypothesis was conducted in the form of a field experiment in three military laboratories. Based on what a sample of 85 scientists and engineers reported on the time they took and the relevance judgments they made with respect to the documents which came across their desks over a four week period, there does not appear to be any evidence to support the assumption that abstracts, accompanying a document, have any significant effect.