How to Loop Through Files in a Directory in Python

You can loop through files in a directory in Python with Path(folder).iterdir() or os.listdir(), and keep only the files. For one file type, glob("*.csv") filters as it loops:

from pathlib import Path

for path in Path("sales_2026").glob("*.csv"):
    print(path.name)

These examples run against a small sales_2026 folder of state sales files that the script rebuilds every time, on Python 3.12.5.

I’ll go through os.listdir and os.scandir, pathlib, filtering by extension, looping through subfolders, and finally reading every CSV to total the sales.

Loop through files in a directory with os.listdir and os.scandir

os.listdir() gives you names only, including folders, so join the path and check isfile(). os.scandir() returns entries that already know their type and size:

import os

folder = "sales_2026"
for name in os.listdir(folder):
    path = os.path.join(folder, name)
    if os.path.isfile(path):             # skip the archive subfolder
        print(name)

print("--- os.scandir with sizes ---")
with os.scandir(folder) as entries:
    for entry in entries:
        if entry.is_file():
            print(f"{entry.name:<12} {entry.stat().st_size:>3} bytes")

Output:

florida.csv
ohio.csv
readme.txt
texas.csv
--- os.scandir with sizes ---
florida.csv   30 bytes
ohio.csv      17 bytes
readme.txt    20 bytes
texas.csv     31 bytes
Command Prompt output of Python os.listdir and os.scandir looping through files in a folder and printing file sizes
Both loops skip the archive folder; scandir also reports each file’s size.

I reach for scandir when I need file details, because it avoids an extra system call per file. Note the order isn’t guaranteed; sort the names if it matters.

Loop through files with pathlib

With pathlib you get full Path objects with handy attributes like .name, .suffix and .stem:

for path in sorted(root.iterdir()):
    kind = "folder" if path.is_dir() else path.suffix
    print(f"{path.name:<12} {kind}")

Output:

archive      folder
florida.csv  .csv
ohio.csv     .csv
readme.txt   .txt
texas.csv    .csv

Paths join with / instead of a function call, which I find easier to read than os.path.join().

Loop through files of one type with glob

If you only want certain files, glob() matches a pattern in one folder. rglob() searches every subfolder too:

print("top-level CSVs:", sorted(p.name for p in root.glob("*.csv")))
print("all CSVs:      ", sorted(str(p.relative_to(root)) for p in root.rglob("*.csv")))

Output:

top-level CSVs: ['florida.csv', 'ohio.csv', 'texas.csv']
all CSVs:       ['archive\\texas_2025.csv', 'florida.csv', 'ohio.csv', 'texas.csv']
Command Prompt output of Python pathlib glob and rglob looping through CSV files in a folder and its subfolders
glob found three CSVs; rglob also found the one inside archive.

To see everything in a folder at once, see listing files in a directory.

How do you loop through files in subfolders?

os.walk() visits every folder in the tree and hands you its path, subfolders and files at each step:

import os

for dirpath, dirnames, filenames in os.walk("sales_2026"):
    print(dirpath, "->", sorted(filenames))

Output:

sales_2026 -> ['florida.csv', 'ohio.csv', 'readme.txt', 'texas.csv']
sales_2026\archive -> ['texas_2025.csv']
Command Prompt output of Python os.walk looping through a folder and its archive subfolder
One line per folder: the top folder first, then archive.

You can prune the walk by editing dirnames in place, for example removing "archive" so it’s skipped.

Process every file in a directory

Here’s the real goal most of the time: open each file and do something with it. This loop reads every state CSV and totals the sales:

import csv

grand_total = 0.0
for path in sorted(root.glob("*.csv")):
    with path.open(newline="", encoding="utf-8") as f:
        total = sum(float(amount) for _city, amount in csv.reader(f))
    grand_total += total
    print(f"{path.stem.title():<8} ${total:>9,.2f}")
print(f"{'Total':<8} ${grand_total:>9,.2f}")

Output:

Florida  $ 2,620.75
Ohio     $   640.00
Texas    $ 2,230.75
Total    $ 5,491.50
Command Prompt output of Python looping through CSV files in a folder, totaling sales per state and a grand total
Three state files read in sorted order, with a grand total of $5,491.50.

Opening each file with with closes it right after it’s read; see Python with open. For cleaning a folder instead, see deleting all files in a directory.

Which way should you loop through files?

NeedUse
Just the names in one folderos.listdir()
Names plus size or type, fastos.scandir()
Path objects, clean codePath.iterdir()
Only one file typePath.glob("*.csv")
Every subfolder tooPath.rglob() or os.walk()

These file and folder guides go well with this one:

Frequently asked questions

How do I loop through files in a directory in Python?

Use for path in Path(folder).iterdir(): and check path.is_file(), or loop over os.listdir(folder).

How do I loop through only CSV files?

Use Path(folder).glob("*.csv"), or rglob to include subfolders.

How do I include subfolders?

Use Path.rglob("*") or os.walk(), which visits every folder in the tree.

Why are my files in a random order?

Directory listings aren’t sorted. Wrap the loop in sorted() for alphabetical order.

What’s faster, os.listdir or os.scandir?

os.scandir is usually faster when you also need file types or sizes, because entries carry that information.

Where is os.scandir documented?

See os.scandir in the Python docs.