Study data came from 12 district courtrooms in 10 circuits. Stenographic reports, as the official court reporters, took the official records and prepared transcripts. Four-track cassette tape recorders were also installed in 11 project courtrooms; an eight-track reel-to-reel recorder was installed in one. Court clerk's office employees operated the recorders, prepared logs, and shipped the recordings to designated transcription companies whenever a transcript was ordered from the official court reporter. The audio-based transcript matched the audiotape in 58 percent of the 5,717 discrepancies, while the steno-based transcript matched it in 42 percent. Five-sixths of the audio-based transcripts were delivered to the clerk of court within 30 days; all were delivered within 35 days. For steno transcripts, the figures were 64 percent and 77 percent. An audio-based system is projected to cost less than half that of a stenographic reporting system for a year. If audio recording becomes an official court reporting method, Federal courts should ensure proper management, reliable selection of transcription service, and adequate operator training. Tabular data, footnotes, and appendixes presenting forms, 67 references, and related materials are supplied.