How to Find Supreme Court Audio Transcripts

The U.S. Supreme Court publishes written transcripts and audio recordings of its oral arguments. This guide shows how to find both for a specific case, how to reach older recordings, and how to check a transcript or make your own from the audio. It covers the U.S. Supreme Court only. Other countries' highest courts publish hearings differently.

Before you start

Collect as much as you can about the case. The most useful details are the case name, the docket number and the date of oral argument. Docket numbers combine the year and a sequence number, such as 22-451. The Court's term begins in October and takes its name from that year, so an argument heard in March 2024 belongs to the October Term 2023. Using the wrong term year is the most common reason a search comes up empty.

You need an internet connection and two kinds of software: a PDF reader for the official transcripts and a media player that can open downloaded audio files. If you plan to produce your own transcript, find out whether your speech-to-text tool runs locally on your computer or sends audio to an online service. The tool's own documentation should say this. Don't assume either way. Argument recordings are public records, but any notes, annotations or other files you upload with them may be kept by the service under its own retention policy.

Step-by-step instructions

Find the official transcript and audio

• Open the Supreme Court's official website and go to its oral arguments section. Menu labels change from time to time, so look for the pages listing argument transcripts and argument audio.
• Choose the term that matches the argument date. Remember that the term is named after the year it began.
• Find the case in the list by argument date, case name or docket number. Consolidated cases are usually listed under the lead case's name and docket number.
• Open or download the transcript PDF. Save it with the docket number in the file name so it is easy to match with the audio later.
• Go to the audio listing for the same term, find the same case and download the recording. Keep the original file unchanged and work on copies.

Find older or synchronized recordings

The Court has posted argument audio on its website since the 2010 term. Recordings from earlier terms were deposited with the National Archives, and academic projects have digitized many of them. The Oyez project, for example, pairs archival audio with transcript text so you can follow along as people speak. For research, a synchronized version is convenient, but cite the official transcript as the record. Recordings of opinion announcements from the bench are separate from argument audio and are not included in the official argument transcripts.

Create your own transcript from the audio

• Check which file formats and sample rates your speech-to-text tool accepts. Convert a copy only if the tool requires it. Upsampling a recording does not add detail.
• Run the tool on the copy. If it offers speaker separation, turn it on, because arguments switch quickly between justices and advocates.
• Older archival recordings often have hiss and background noise. Process the file once with noise reduction and once without, then keep whichever result reads more accurately. Heavy filtering can make speech recognition worse.
• If you have no choice but to record playback instead of downloading the file, select a system-audio or loopback input device rather than the microphone. Set the input level so the peaks never clip, and mute other apps that play sound through the same output device.

Check the output

Compare the transcript's cover page with the details you collected: case name, docket number and argument date. Then play a few points from the beginning, middle and end of the audio and check that the transcript wording matches. The official transcript is prepared by a court reporter and is the version normally cited. It can still contain small errors, so if exact wording matters, confirm it against the recording.

Machine-made transcripts need closer review. Check speaker attribution, case names, statute citations, Latin phrases and the names of justices and counsel, because speech recognition often gets these wrong. Make sure the transcript covers the full length of the recording, since some tools quietly skip long files or stretches of silence.

Common problems and next steps

• The case is missing from the list: confirm the term year and docket number. Some cases are decided without oral argument, so no transcript or audio exists for them.
• The transcript is available but the audio is not: audio and transcripts can be released on different schedules. Check the audio listing again later.
• The justices are not named in an older transcript: earlier transcripts often labeled questions from the bench only as QUESTION. Identify the speaker from the audio or from a synchronized archival version.
• The audio file won't play: download it again in case the transfer was interrupted, or try another media player before converting the file.
• The machine transcript mixes up speakers: the tool may be struggling with overlapping speech. Correct the speaker labels by hand against the audio rather than running the tool again on the same file.