CelesTrak Usage Policy

by Dr. T.S. Kelso
2026 May 15
Updated 2026 May 22

Users need to comply with our guidelines to minimize overuse of CelesTrak, as described at:

https://celestrak.org/NORAD/documentation/gp-data-formats.php (for GP data)

There is a lot of detail here, including what we have had to do to keep CelesTrak functional under recent dramatic increases in usage.

Here’s what we need you to do:

• Ensure you are using queries inline with our documentation:
• GP (https://celestrak.org/NORAD/documentation/gp-data-formats.php)
• SupGP (https://celestrak.org/NORAD/documentation/sup-gp-queries.php)
• SATCAT (https://celestrak.org/satcat/satcat-format.php)
• SOCRATES Plus (https://celestrak.org/SOCRATES/socrates-format.php)
• Only download the data you need, when you are going to use it, and only download data once per update.
• For GP data, updates are once every 2 hours.
• For SupGP data, which updates on different schedules for different constellations, please use 2 hours, as well.
• EOP data updates once per day.
• Space weather data updates every 3 hours.
• The SATCAT updates manually once or twice a day.
• SOCRATES Plus updates every 10-11 hours (right now).
• GPS data typically only updates once per day.
• Many data sets (e.g., SATCAT, EOP and space weather, or GPS data) have queries to get directories to check for the latest data, such as:
  • SATCAT: https://celestrak.org/satcat/jsonDir.php
  • EOP & Space Weather: https://celestrak.org/SpaceData/jsonDir.php
  • SOCRATES Plus: https://celestrak.org/SOCRATES/jsonDir.php
  • GPS Almanacs & Ops Advsory: https://celestrak.org/GPS/jsonDir.php
  • GPS NANU: https://celestrak.org/GPS/NANU/jsonDir.php
  They show the same information as on the parent page. Please don’t query these more than once per hour.
• Note that we are starting to enforce some of those limits for larger data sets, like the Active list for GP data and the Starlink list for GP & SupGP data. The same will be true for SATCAT, SOCRATES, EOP, and space weather soon. Users will be encouraged not to download 60 years of EOP and space weather every time, as well.
• Most importantly, we send HTTP error responses when users are exceeding limits or using incorrect (long-outdated) URLs (e.g., HTTP 301, 403, 404, 50x). M2M (machine-to-machine) software should immediately stop querying when it receives any non-HTTP 200 responses and report the results to a human for investigation. Repeatedly ignoring them will end up sending your IP address to the firewall.
  This step alone will prevent your IP address from being sent to the firewall, even if we have to make further changes.
• HTTP 301 means the URL is being redirected because it is either going to the wrong domain (i.e., something other than https://celestrak.org) or using legacy queries that we have been redirecting for years now, but will soon stop.
• HTTP 404s are also signs of using old URLs—usually because you are using some software written a decade or more ago.
• HTTP 50x responses are server errors, meaning the server is struggling under heavy load. Queries need to stop immediately to allow system recovery for all users.
• Please be sure your software developer(s) is(are) aware of these problems and our usage policy, so that they can update their software for all of their users.
• Much of this is being done because the legacy formats of the SATCAT, GP, and SupGP data will change in just a copule of months (currently predicted for mid-July) when we run out of 5-digit catalog numbers at 69999 not 99999. Users need to update their software to use the CSV formats, which have been available for 6 years now. The CSV format has many advantages, including ease of use in a wide variety of applications, and is even smaller (~10% smaller) than the TLE format and far easier to parse.

Please realize that CelesTrak supports hundreds of thousands of unique IP addresses each day and millions each month. We need to ensure all users respect our resource limitations in order to be able to continue to provide reliable service.

And while we realize that many companies or organizations have chosen to set up NATs to have all of their users coming from the same IP address, instead of implementing IPv6 (which has been an Internet standard since 2017), we have no way to know how many users are accessing CelesTrak via a single IP address. It is up to those organizations to figure out a way to set up a proxy to cache these queries in a way that doesn’t get all of their employees or devices blocked, by caching requests on their end.

Bottom line: There is no way for CelesTrak to set up a system to manage millions of users without having to charge for access, which would immediately break our long tradition of making data freely available to all users. We just need users to work with us to ensure that CelesTrak continues to be a reliable service and educational platform for the community. I appreciate your understanding and support and encourage you to spread the word. - TS