Environment Variables in Python: Read and Set

A Python application can use different settings in development and deployment without changing its source code. Environment variables carry those settings into a program when it starts, and Python lets your code read or change them.

I’ll show you how to choose what happens when a setting is missing and how to read or set values in Python.

TL;DR

Python reads environment values through the os module, and os.environ lets your program read or change values for its own process.

• Use os.getenv() with a fallback when a setting is optional.
• Use os.environ[“NAME”] when a missing setting should raise KeyError.
• Convert values such as ports and Boolean flags from strings before using them.
• Python changes do not rewrite the shell that launched it.

What Are Environment Variables in Python?

An environment variable is a named text value supplied to a process by its launching environment. Applications use these values for settings that can vary across the environments where code runs, without editing the source.

Python exposes environment settings through os.environ, a dictionary-like mapping captured when os is first imported.

In the table, NAME stands for the exact key text. Replace it with a value such as APP_PORT, keeping the spelling consistent where the setting is supplied and read.

For the same bracket and get() syntax with ordinary mappings, see AskPython’s dictionary tutorial.

How to Read and Set Environment Variables in Python

The example uses only Python’s standard library, so you can run it without installing a package or editing the shell.

Step 1: Choose how a missing key should behave

For a required key, let KeyError stop startup or catch it at the configuration boundary to add context. If you catch it, include the missing key’s name. For an optional region, os.getenv(“REGION”, “local”) gives the program a clear local-development choice.

Step 2: Convert text values before using them

Environment values arrive as strings, so convert numbers with int() or float() and parse flags with an explicit rule, since bool(“false”) is True for a non-empty string.

Conversion can fail too. Catch ValueError at the configuration boundary and report the variable name and expected type, but do not include a credential or its value in an error message.

Step 3: Set a value and run the example

This example assigns sample values for lookup, fallback, conversion, and update, so it prints only its own values. I ran environment_variables_demo.py as shown in the terminal image. I saw all six output lines in that capture.

import os

# These values live only in this Python process.
os.environ["APP_MODE"] = "staging"
os.environ["APP_PORT"] = "8000"
os.environ["DEBUG"] = "false"

mode = os.getenv("APP_MODE")
missing_region = os.getenv("DEPLOY_REGION")
default_region = os.getenv("DEPLOY_REGION", "local")
port = int(os.environ["APP_PORT"])
debug = os.getenv("DEBUG", "false").lower() in {"1", "true", "yes", "on"}

print(f"mode={mode}")
print(f"missing region={missing_region}")
print(f"default region={default_region}")
print(f"port={port} ({type(port).__name__})")
print(f"debug={debug}")

os.environ["APP_MODE"] = "production"
print(f"mode after update={os.environ['APP_MODE']}")

Run this exact command:

python3 environment_variables_demo.py

The captured output is:

mode=staging
missing region=None
default region=local
port=8000 (int)
debug=False
mode after update=production

The missing lookup prints None, while the second lookup supplies local. The port is converted from text to an integer, and the accepted-string rule turns false into the Boolean False. The final line shows that assigning the same key again replaces its value in this process.

Use os.environ.items() when you need to iterate over key/value pairs. AskPython’s Python for loop guide explains the loop syntax. Avoid printing the entire mapping because it can contain credentials or other private configuration.

Common Errors: Missing Values, Types, and .env Files

Configuration failures often come from a missing name, a value in the wrong format, or a mistaken expectation about where a change applies. Check these cases before debugging the code that consumes the setting.

A Python assignment changes os.environ for the running process and is inherited by later subprocesses, but it does not rewrite the shell profile or affect unrelated processes.

Check empty values and platform behavior

Settings can look similar in a launch configuration yet behave differently in Python.

Use .env files for local development

Python does not load .env files by itself. The python-dotenv documentation describes how to opt in to loading one.

Conclusion: Keep Configuration at the Process Boundary

Use os.environ or os.getenv() to read values, choose strict failure or a fallback deliberately, and convert text before the application relies on it. Keep the distinction between changing one Python process and configuring the process that launches it.

For exact API behavior, see the Python documentation for os.environ and os.getenv. If you meant Python’s executable search path rather than application settings, see AskPython’s guide to PYTHONPATH in the shell.

FAQ

These answers cover nearby setup questions that can be confused with reading application configuration.