How to Close a File in Python

To close a file in Python, call close() on the file object, or use a with block, which closes it for you even after an error:

with open("orders.txt", encoding="utf-8") as f:
    data = f.read()
print(f.closed)   # True

My machine for these runs is Python 3.12.5 on Windows 11, which matters for the Windows lock section.

In this tutorial, we’ll look at close(), the closed attribute, flush versus close, try/finally, and what goes wrong when you leave a file open.

Close a file with close() and check the closed attribute

close() writes out buffered data and frees the OS handle. Opening modes are covered in opening a file in Python.

f = open("orders.txt", "w", encoding="utf-8")
print("closed before:", f.closed)
f.write("1001,Emily Carter,49.99\n")
f.close()
print("closed after: ", f.closed)

f.close()                      # a second close() does nothing

try:
    f.write("1002,James Lee,15.00\n")
except ValueError as err:
    print("ValueError:", err)

Output:

closed before: False
closed after:  True
ValueError: I/O operation on closed file.

If you ever see “I/O operation on closed file”, a read or write sits outside the block that owned the file.

What’s the difference between flush() and close()?

Writes land in a memory buffer first. flush() pushes it to disk and keeps the file open; close() flushes, then shuts it. A second reader checks the file between steps:

log = open("sales_log.txt", "w", encoding="utf-8")
log.write("2026-03-02,Emily Carter,129.50\n")

with open("sales_log.txt", encoding="utf-8") as reader:
    print("before flush:", repr(reader.read()))

log.flush()
with open("sales_log.txt", encoding="utf-8") as reader:
    print("after flush: ", repr(reader.read()))

log.write("2026-03-03,James Lee,88.25\n")
log.close()                    # close() flushes the rest first
with open("sales_log.txt", encoding="utf-8") as reader:
    print("after close: ", reader.read().count("\n"), "lines on disk")

Output:

before flush: ''
after flush:  '2026-03-02,Emily Carter,129.50\n'
after close:  2 lines on disk

That’s why a log can look empty while your script runs. You’ll want flush() in long-running writers. More on writing text is in writing a string to a file in Python.

Close a file safely with try/finally

If an exception fires before close(), that line never runs, but finally always does. Here the second order’s amount is N/A:

with open("orders.txt", "w", encoding="utf-8") as out:
    out.write("1001,Emily Carter,49.99\n1002,James Lee,N/A\n")

def load_orders(path):
    f = open(path, encoding="utf-8")
    try:
        return [float(line.split(",")[2]) for line in f]
    finally:
        f.close()
        print("finally ran, closed =", f.closed)

try:
    load_orders("orders.txt")
except ValueError as err:
    print("caught:", err)

Output:

finally ran, closed = True
caught: could not convert string to float: 'N/A\n'

The file was closed before the caller caught the error. For several error types, see catching multiple exceptions in Python.

Let the with statement close files automatically

A with block is that same try/finally, written for you:

try:
    with open("orders.txt", "w", encoding="utf-8") as f:
        f.write("1001,Emily Carter,49.99\n")
        raise RuntimeError("payment API timed out")
except RuntimeError as err:
    print("error:", err, "| closed:", f.closed)

Output:

error: payment API timed out | closed: True

The file closed even though the block raised. I use with for nearly every file; the full guide is Python with open.

What happens if you don’t close a file?

CPython usually closes a forgotten file during garbage collection. It’s still a bug, and Python flags it as a ResourceWarning once you turn warnings on:

import warnings
warnings.simplefilter("default")   # ResourceWarning is hidden unless you turn it on

with open("orders.txt", "w", encoding="utf-8") as f:
    f.write("1001,Emily Carter,49.99\n1002,James Lee,15.00\n")

def count_orders(path):
    f = open(path, encoding="utf-8")   # opened, never closed
    return len(f.readlines())

print("orders:", count_orders("orders.txt"))

Output:

C:\pyguides\unclosed_warning.py:11: ResourceWarning: unclosed file <_io.TextIOWrapper name='orders.txt' mode='r' encoding='utf-8'>
  print("orders:", count_orders("orders.txt"))
ResourceWarning: Enable tracemalloc to get the object allocation traceback
orders: 2

A crash can lose unflushed data. Run tests with python -W default to catch leaks. Other file object methods are in Python file methods.

Why can’t Windows delete a file I didn’t close?

Windows locks a file while it’s open, so deleting or renaming it fails with WinError 32. Linux and macOS allow it, so the bug hides until you run on Windows:

import os

f = open("report_2026.csv", "w", encoding="utf-8")
f.write("region,total\nEast,1250.00\n")

try:
    os.remove("report_2026.csv")
except PermissionError as err:
    print("PermissionError:", err)

f.close()
os.remove("report_2026.csv")
print("deleted after close:", not os.path.exists("report_2026.csv"))

Output:

PermissionError: [WinError 32] The process cannot access the file because it is being used by another process: 'report_2026.csv'
deleted after close: True

With WinError 32, hunt for the open handle, sometimes in Excel. Safe delete patterns are in deleting a file if it exists in Python.

Next on the file-handling list:

• Read a file line by line in Python
• Write multiple lines to a file in Python
• Check if a file exists in Python
• Create a new file in Python

Frequently asked questions