Here's an article from ChessBase on how to do it: ChessBase 18 – Tips for Beginners, Part 7: Organising databases! by Stefan Liebig 2025.03.20 (via the wayback machine because chessbase.com is missing the images) 
https://web.archive.org/web/20250425092524/https://en.chessbase.com/post/chessbase-18-tips-for-beginners-part-7-organising-databases

It's a nightmare.

I watched the recent video How Grandmasters Organize Their Opening Files in ChessBase by GM Felix Blohberger 2026.02.15 https://youtu.be/F6aZT-xRyj4 , and I was struck by three related thoughts:

1. Blohberger became a GM doing it the "wrong" way, which goes to show how important it is to do things the "right" way. His old way seems in fact very similar to what Liebig presents. In a recent training camp by Anand, the revelation was -- instead of keeping each opening in its own database, keep them all in the same database.
2. Blohberger is arguably still doing it the "wrong" way, because he went from multiple database files to two. But by all means watch the video to see how he cleverly hijacks the ChessBase buttons for White Repertoire and Black Repertoire.
3. Clever, but this should not be necessary! For a database purist, there should be only one database and two views. But ChessBase doesn't do views.

Well we all had ideas, and they seemed good when we came up with them. Until we decided another way was better. So let's hear some current and abandoned ways of organizing files. For the moment I will list a few good reasons for having multiple databases, even though I try hard to use just one database, Anand-style.

• Avoid over-sharing, e.g. in a student+coach or player+second relationship the other party doesn't get to see all the data, but only the data relevant to the relationship.
• Working files for a book or course need to be in their own database.
• Files from an external source should be segregated and sanitized before being imported.

And here is an organizational structure I came up with to mitigate the "two computers" problem, aka the dreaded "conflicted copy". On any internal hard drive, external usb drive, or cloud storage service, I create three (sometimes four) folders. Note that my GNU/Linux operating system has a ~/Documents folder, and I create these three (four) folders inside the Documents folder. Other folders like ~/Music, ~/Pictures, ~/Videos I just ignore. So I have some chess videos but they also go inside Documents.

1. canonical
   - Files where the primary copy exists in this folder. Although each device has a canonical folder, each file (usually by subject) can have only one canonical location. So the canonical location for my scripts is Dropbox, for my passwords it's the encrypted laptop hard drive, for my address book it's gmail, for my chess archives it's a usb drive, for my working chess database it's Dropbox again, and so on.
2. dynamic
   - Files which have been (a) created on this device or (b) copied from the canonical location for editing. Either way, they are destined at some future time to be copied to the canonical location, and once copied there they are deleted from the dynamic location (although if I am feeling especially squirrelly I move them to static instead of deleting). A chess example would be some annotated game from chessbase.com.
3. static
   - Files which have been copied from the canonical location for reference. I treat these as read-only. If anything about these files were to change (very unusual for static files), they should be relocated to dynamic to reflect that status. A chess example would be Tim Harding's Ultra Corr database.
4. stage
   - A special kind of dynamic location which is updated by a script and rsynced to a removable drive. The idea is I can run the script from two (or more) different laptops, whichever one I happen to have with me. Because they each rsync with the same external drive no update is duplicated. Due to my paranoia about scripted updates, rsync doesn't delete, and the removable drive is a stage location _not_ a canonical location. A chess example would be TWIC downloads.

All of these folders are completely independent of backups, which are nothing to do with organization. But backups are so important that I mention them a lot. And remember, cloud storage is not a backup!