How to Reinstall Python Without Losing Projects, Packages or Virtual Environments

Preserve first

Protect projects before touching Python

Normal Python uninstall/reinstall operations target the interpreter, not source folders you created elsewhere. The real risk is losing track of dependencies or breaking virtual environments that reference the old interpreter.

GoalKeep source code untouched

Before uninstallRecord version + path

RestoreRebuild the venv

Record the active setup

python --version
python -c "import sys; print(sys.executable)"
python -m pip --version

Export dependencies

python -m pip freeze > requirements-backup.txt

If the project already uses requirements.txt, pyproject.toml, Poetry, Pipenv, uv, Conda, or another lock file, preserve that project metadata as the preferred dependency record.

Virtual environment

Do not treat the old .venv as the backup

A virtual environment points to a specific base Python installation. After the base interpreter is removed or moved, that environment can stop launching. Keep the source project and dependency metadata; rebuild the environment after reinstalling Python.

Clean removal

Remove only the Python installation you intend to replace

Open Windows Installed apps and identify the version tied to the broken interpreter path you recorded. Do not delete every Python entry simply because multiple versions appear. Side-by-side versions are supported and may belong to different projects.

If you use the current Python Install Manager, the manager itself and the Python runtimes it manages are separate pieces.

Reinstall

Install the required Python version again

For current Windows releases, Python recommends the Python Install Manager. Install a normal stable runtime with:

py install

If a project requires a particular release, request that release explicitly.

py install 3.13

Older Python releases may use the classic Windows executable installer from Python.org. Match the version and architecture your project expects instead of automatically moving to the newest release.

Project restore

Create a fresh environment and restore dependencies

python -m venv .venv
.venv\Scripts\activate
python -m pip install -r requirements-backup.txt

Then reopen the project in VS Code or another IDE and select the new .venv interpreter. Source files, Git history, notebooks, assets and databases stored in the project directory remain independent from the Python application.

Final checkpoint

Verify Windows is using the new interpreter

python --version
where.exe python
python -c "import sys; print(sys.executable)"
python -m pip --version

If the path or version is still wrong, stop reinstalling. The remaining issue is usually PATH, an app execution alias, a launcher conflict, or IDE interpreter selection.

Project inventory

Back up what Python does not know how to recreate

Source code, notebooks, templates, static assets and local data files that are not already committed to version control.
Dependency files such as requirements.txt, pyproject.toml, lock files or Conda environment definitions.
Local databases, uploaded media and generated files stored outside the project’s normal backup process.
A private record of environment-variable names and where their secret values are stored. Do not commit API keys or copy them into a public requirements file.

Test the backup by opening a few important files before uninstalling anything. Then recreate the virtual environment from dependency metadata. Python’s documentation treats virtual environments as disposable and not safely movable or copyable.

After restoration, run the project’s real entry point or test suite—not only python --version. A healthy interpreter can still be missing a dependency, native build tool or IDE interpreter selection.

FAQ checkpoint

Frequently asked questions

Will reinstalling Python delete my .py files?

No, not when your projects are stored in normal project folders. The main compatibility risk is the interpreter and virtual environment, not the source files.

Should I copy site-packages before reinstalling?

Prefer exporting dependencies or using the project lock/configuration files. Copying site-packages can carry incompatible binaries and stale paths.

Can I keep the same virtual environment?

Sometimes, but recreating it is usually safer after replacing the base interpreter.

Next mission

Related Python fixes

Continue with the page that matches the symptom you see after this step.

Repair a broken Python installation →Windows runs the wrong Python version →

Reference log

References

Installation behavior and commands were checked against current Python documentation on 10 August 2026.