How to Make a Directory in Python (os and pathlib)

Need to make a directory in Python? Use os.mkdir("name") for a single folder, or Path("a/b/c").mkdir(parents=True, exist_ok=True) to create nested folders without errors if they already exist:

from pathlib import Path

Path("exports/2026/q4").mkdir(parents=True, exist_ok=True)

Each folder in these examples was created for real by Python 3.12.5 on Windows, and the screenshots show the Command Prompt output.

I’ll cover os.mkdir(), os.makedirs(), pathlib, dated and per-customer folders, temporary directories, and the errors you’ll hit along the way.

How to make a directory in Python with os.mkdir()

os.mkdir() creates exactly one folder. It raises an error if the folder already exists, or if its parent folder is missing:

import os

os.mkdir("reports")
print("created:", os.path.isdir("reports"))

try:
    os.mkdir("reports")              # second time: it already exists
except FileExistsError as err:
    print("FileExistsError:", err)

try:
    os.mkdir("data/2026/october")    # parent folders don't exist yet
except FileNotFoundError as err:
    print("FileNotFoundError:", err)

Output:

created: True
FileExistsError: [WinError 183] Cannot create a file when that file already exists: 'reports'
FileNotFoundError: [WinError 3] The system cannot find the path specified: 'data/2026/october'

The [WinError 183] and [WinError 3] codes are Windows’ own error numbers. On macOS and Linux you’ll see [Errno 17] and [Errno 2] with the same meaning.

Creating nested folders with os.makedirs()

os.makedirs() creates every missing level of a path at once. Add exist_ok=True and it quietly does nothing when the folders are already there:

import os

os.makedirs("invoices/2026/10", exist_ok=True)     # builds every missing level
os.makedirs("invoices/2026/10", exist_ok=True)     # no error the second time

for root, dirs, files in os.walk("invoices"):
    print(root)

Output:

invoices
invoices\2026
invoices\2026\10

That makes it safe to call at the top of any script that writes output, without first checking whether the folder exists.

Making a directory with pathlib (the modern way)

pathlib treats paths as objects, so you join them with / and call methods on them. mkdir(parents=True, exist_ok=True) matches os.makedirs(..., exist_ok=True):

from pathlib import Path

out = Path("exports") / "2026" / "q4"
out.mkdir(parents=True, exist_ok=True)

(out / "sales.csv").write_text("city,total\nAustin,1250.50\n", encoding="utf-8")

print("folder:", out, "| exists:", out.is_dir())
print("contents:", [p.name for p in out.iterdir()])
print("absolute path ends with:", out.resolve().parts[-3:])

Output:

folder: exports\2026\q4 | exists: True
contents: ['sales.csv']
absolute path ends with: ('exports', '2026', 'q4')

I use pathlib in all new code, since it reads better and behaves the same on every OS. Need to know where you are first? See getting the current directory in Python.

Creating a dated directory and subfolders

A common job is one folder per day, or one per store or customer. f-strings with date format codes build the names, and a loop handles the rest:

from datetime import date
from pathlib import Path

today = date(2026, 10, 22)
folder = Path("backups") / f"{today:%Y-%m-%d}"
folder.mkdir(parents=True, exist_ok=True)
print("backup folder:", folder.as_posix())

stores = ["Austin", "Denver", "Boston"]
for store in stores:
    (Path("orders") / store.lower()).mkdir(parents=True, exist_ok=True)
print(sorted(p.name for p in Path("orders").iterdir()))

Output:

backup folder: backups/2026-10-22
['austin', 'boston', 'denver']

Making a temporary directory in Python

For scratch work, tempfile.TemporaryDirectory() makes a folder in your system’s temp location and deletes it, with everything inside, when the with block ends:

import tempfile
from pathlib import Path

with tempfile.TemporaryDirectory() as tmp:
    work = Path(tmp)
    (work / "draft.txt").write_text("temporary notes", encoding="utf-8")
    print("inside: exists?", work.exists(), "| files:", [p.name for p in work.iterdir()])

print("after:  exists?", work.exists())     # deleted automatically

Output:

inside: exists? True | files: ['draft.txt']
after:  exists? False

It’s great for tests and for processing downloads you don’t want to keep. The with pattern is the same one explained in Python with open() as file.

Why does exist_ok still raise FileExistsError?

exist_ok=True only ignores an existing folder. If a regular file already has that name, Python still raises FileExistsError, because it can’t turn a file into a folder:

from pathlib import Path

target = Path("photos")
print("before:", target.exists())

if not target.exists():
    target.mkdir()
print("after: ", target.is_dir())

clash = Path("photos.txt")
clash.write_text("not a folder", encoding="utf-8")
try:
    clash.mkdir(exist_ok=True)       # exist_ok only ignores an existing *folder*
except FileExistsError as err:
    print("FileExistsError:", err.strerror)
clash.unlink()

Output:

before: False
after:  True
FileExistsError: Cannot create a file when that file already exists

Checking with exists() first is fine for scripts, but exist_ok=True avoids a race if two processes run at once. For more checks, see checking if a directory exists in Python.

os.mkdir vs os.makedirs vs Path.mkdir

Function	Creates parents?	Ignore if it exists?	Best for
os.mkdir(p)	No	No	One folder, and you want an error if it’s there
os.makedirs(p, exist_ok=True)	Yes	Yes	Older code using os
Path(p).mkdir(parents=True, exist_ok=True)	Yes	Yes	New code
tempfile.TemporaryDirectory()	n/a	n/a	Scratch space that deletes itself

These folder and file guides go well with this one:

• List files in a directory with Python
• Copy files to another directory
• Copy a file in Python
• Check if a file exists in Python

Frequently asked questions