My Python project won't install its dependencies. How do I fix it?

Also searched as: pip install error externally-managed-environment · pip install error · Python dependency conflicts · pip install fails for every package: Could not find a version that satisfies the requirement · Pip installation in virtualenv broken

Updated 3 October 2026

Short answer

Create a fresh virtual environment using the Python version the project was built for, then install from its requirements file. Most pip failures on old projects are mismatches: a Python too new for the project's pinned packages, packages that conflict with each other, or a system Python that refuses installs outside a virtual environment, which is what the "externally-managed-environment" error means.

What's the problem

pip install -r requirements.txt fails. You see "externally-managed-environment", "Could not find a version that satisfies the requirement", "ResolutionImpossible", "metadata-generation-failed", or a compiler error. The project used to install, on someone else's machine, years ago.

Why it happens

  • The system Python is protected. Many operating systems now mark their own Python as externally managed, so pip refuses to install into it and points you to a virtual environment.
  • Your Python is too new, or too old. Packages pinned years ago may not support today's Python, so pip finds no matching version.
  • The pins conflict. Two packages require incompatible versions of a third, and pip's resolver can't satisfy both.
  • Packages need compiling. Older versions of scientific, image or database packages build from source when no prebuilt version matches your Python, and fail without the right compilers and system libraries.
  • The old environment was half-broken. A virtual environment copied between machines or Python versions doesn't work. Environments are rebuilt, not moved.

How to fix it

  1. Find the Python version the project expects. Check runtime.txt, .python-version, pyproject.toml (requires-python), the Dockerfile and any build settings.
  2. Install that version with your system's tools or a version manager such as pyenv, alongside your current Python.
  3. Create a fresh virtual environment with it: python -m venv .venv, then activate it.
  4. Install from the requirements file: python -m pip install -r requirements.txt.
  5. If pip can't find a version, the pinned version doesn't support this Python, or the package was renamed. Check the package's page on PyPI.
  6. If it reports conflicts, read which packages disagree and loosen or update one pin at a time.
  7. Once it installs and runs, record exact versions for every dependency (a lockfile, for example with pip-tools), then upgrade Python and packages in steps.

When to call Preventionlabs

If the right Python and a fresh environment get it installing, you're through the hardest part. Call us when the pinned packages no longer exist for any supported Python, the conflicts chain, or the code depends on library behaviour that's since changed. A resurrection delivers an app that builds and starts on maintained dependencies, deployed in your own hosting account.

Submit your project for a free assessment

Free assessment. $10,000 AUD flat to get it live, only if we take it on and you go ahead.

Sources

  1. Python: venv: Creation of virtual environmentsofficial docsA virtual environment is created on top of an existing Python installation, known as the virtual environment’s “base” Python, and by default is isolated from the packages in the base environment, so that only those explicitly installed in the virtual environment are available.
  2. Python Packaging Authority: Externally Managed Environmentspackaging specificationThis specification defines an EXTERNALLY-MANAGED marker file that allows a Python installation to indicate to Python-specific tools such as pip that they neither install nor remove packages into the interpreter’s default installation environment, and should instead guide the end user towards using Python Virtual Environments.
  3. pip: Dependency Resolutionofficial docsDuring deployment, you can create a lockfile stating the exact package and version number for each dependency of that package.
  4. pip: User Guide: Requirements Filesofficial docs“Requirements files” are files containing a list of items to be installed using pip install like so:

Related

My Docker build fails. How do I fix it? Find the Dockerfile step that failed. Old projects usually fail on an outdated base image, a file outside the build context, or installing dependencies. Everything in my project is years out of date. How do I update it? Get it running exactly as it is first. Then update in small steps (runtime, then framework, then the rest), testing between each, and replace what's abandoned. npm install fails on my old project. How do I fix it? Usually Node is too new for the project, the lockfile and package.json disagree, or a package needs build tools or no longer exists. Check each, in this order. A database migration failed in production. What now? Stop deploying, back up, and find out exactly which steps ran. Then either roll back and fix, or finish the steps by hand and mark it resolved. Never reset.
← All of It used to work