I'm looking for the difference between cp -r and cp -a. What does "recursive" mean in terms of copying files from a folder?

  • 76
    some people are looking for a better explanation to the man pages because it is not always clear what the description meant. I have had to dig many times and test/experiment to figure out what is really going on with a particular command. grep's man page can confuse me any day.
    – Joe
    Aug 7, 2012 at 22:41
  • 2
    I can understand your frustration, but imo, most times best way to learn is to test and experiment.
    – tripledes
    Aug 8, 2012 at 1:09
  • 50
    OR learn by asking questions here, getting the answer in a way you understand, and then have a better sense of understanding what and how to test and experiment. Please keep in mind that works for you very often doesn't work for me, etc. etc.
    – Will Lanni
    Mar 27, 2015 at 0:41
  • 2
    @tripledes, my computational biology professor said that is a great way to somewhat understand the specific cases you've tested without actually understanding the underlying rules. As a scientist, I can tell you that experiments only answer the questions you know to ask.
    – Josh
    Apr 6, 2020 at 15:24

3 Answers 3


Recursive means that cp copies the contents of directories, and if a directory has subdirectories they are copied (recursively) too. Without -R, the cp command skips directories. -r is identical with -R on Linux, it differs in some edge cases on some other unix variants.

By default, cp creates a new file which has the same content as the old file, and the same permissions but restricted by the umask; the copy is dated from the time of the copy, and belongs to the user doing the copy. With the -p option, the copy has the same modification time, the same access time, and the same permissions as the original. It also has the same owner and group as the original, if the user doing the copy has the permission to create such files.

The -a option means -R and -p, plus a few other preservation options. It attempts to make a copy that's as close to the original as possible: same directory tree, same file types, same contents, same metadata (times, permissions, extended attributes, etc.).

  • 54
    Basically, unless you want something special, you never need -r because -a (for archive) is always the safest and probably what you expected to happen.
    – ams
    Aug 8, 2012 at 9:20
  • 3
    @ams Yes, that's a good summary. The only common reason to use -r is because you're on some unix variant other than Linux that doesn't have -a, and generally you'd use cp -rp. Or rsync -a. Aug 8, 2012 at 9:51
  • 1
    sometimes permissions will allow -r but not -a
    – DrCord
    Dec 19, 2015 at 15:40
  • Not -r, not -R doesn't work in Ubuntu 18
    – Arkady
    Jun 17, 2019 at 15:06
  • Good summary, but according to my man page, -a (short for --archive) means -dR --preserve=all. Which is almost the same as what you said. Nov 13, 2021 at 17:39

The -r or -R option for "recursive" means that it will copy all of the files including the files inside of subfolders.

The -a option as listed is the same as -dR which means it will preserve links as well as copy the contents of subdirectories. What it means by preserving links is that it will not follow links as it is recursively copying.

-a=-dR --preserve=all
-d=--no-dereference --preserve=links


-a=-R --no-dereference --preserve=links --preserve=all

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.