Open the Python Packages tool window, type pandas in the search field and click Install. PyCharm puts the package into the interpreter currently selected for the project, using pip by default.
The Terminal is where this goes wrong. It installs into whichever Python the shell finds first, so pandas can install cleanly and still fail on import.
Check the Python version before anything else. PyPI lists pandas 3.0.6, released September 17, 2026, as requiring Python 3.11 or newer, and a project interpreter on Python 3.10 or older sits below that floor.
How to install pandas from the PyCharm interface
The Python Packages tool window and the interpreter settings dialog both install into the interpreter currently selected for the project, using pip by default (PyCharm 2026.2 Help). Conda environments behave a little differently, and that comes up further down.
Python Packages tool window
- Click Python Packages on the left toolbar, or go to View | Tool Windows | Python Packages.
- Look at the interpreter selector in the status bar. It has to be the one your project runs with.
- Type pandas in the search field.
- Select pandas and click Install next to the name for the latest version. For an older release, pick a version from the list first.
- Wait until pandas shows up among the installed packages.
Right-click next to the package name and choose Install with Options if you need pip flags.
None of this is specific to pandas. Any library goes in the same way, as this guide to installing packages in PyCharm shows.
Python interpreter settings
This route gets you the same result, with a version picker added.
- Open Settings | Python | Interpreter, or choose Interpreter Settings in the status bar selector, then pick the project interpreter.
- Click Install on the toolbar and search for pandas in the Available Packages dialog.
- Tick Specify version if you want a release other than the latest.
- Select pandas and click Install Package.
An empty interpreter list means the project has no Python attached yet, and in that case adding a Python interpreter in PyCharm comes before any install.
Conda environments
The pandas documentation (3.0.6) points conda users to conda-forge. pip users get PyPI, and anyone building from source goes through GitHub.
When the selected interpreter is a Conda environment, PyCharm shows a Use Conda Package Manager toggle, enabled by default. The tool window search then returns conda repository packages too (PyCharm 2026.2 Help).
From the command line, the equivalent is conda install -c conda-forge pandas.
One thing to keep in mind: pandas from the Anaconda distribution is not managed by the pandas development team, per the same documentation.
Installing pandas with pip in the PyCharm Terminal
Run pip install pandas in the Terminal tool window and pip downloads pandas from PyPI into whichever Python the shell resolves. That last clause is where terminal installs go wrong.
PyCharm’s troubleshooting page for package installation gives the fix. Activate the project’s virtual environment, or call the interpreter by its full path with -m pip, before running pip, so the package lands in the interpreter you meant. The IDE routes avoid the slip because the project interpreter is already selected (PyCharm 2026.2 Help).
New to the built-in shell? This walkthrough on opening the terminal in PyCharm covers the basics.
A safer command is python -m pip install pandas, which ties pip to the exact python the shell runs.
If you want to see the target before installing, python -c "import sys; print(sys.executable)" prints the interpreter path. It should match the one under Settings | Python | Interpreter.
The pandas 2.0 documentation adds a Linux and macOS tip, which is that which python shows which installation the shell is using.
pandas recommends installing from a virtual environment (pandas documentation, 3.0.6). A project with none invites these mix-ups, and creating a virtual environment in PyCharm takes one dialog.
I’d keep the Terminal as the backup option. The tool window already knows the project interpreter. The shell has to be told.
Which installation method fits your project?
For one package in one project, use the Python Packages tool window. Move to the Terminal, requirements.txt or the conda toggle only when the project calls for it.
| Method | Installs into | Best for | Watch out for |
|---|---|---|---|
| Python Packages tool window | Selected project interpreter | Quick single installs | Wrong interpreter selected |
| Interpreter settings | Interpreter chosen in the dialog | Picking a version, upgrading | Docker-based interpreters |
| Terminal with pip | Python the shell resolves | Repeatable commands | Environment not activated |
| requirements.txt | Interpreter PyCharm sets up or selects | Shared projects | pandas missing from the file |
Create the environment first and pick the method after. Which environment type to create is its own decision, and a look at how venv and Conda environments compare helps before you commit.
With requirements.txt, PyCharm prompts you to install the packages listed in the file once it is added to the project (PyCharm 2026.2 Help).
When pip-based methods do not apply
Pipenv projects work differently. PyCharm installs, removes and updates packages through pipenv rather than pip, with sources defined in the Pipfile (PyCharm 2026.2 Help).
Docker-based interpreters are excluded from PyCharm’s install, uninstall and upgrade options for packages, so pandas goes into the image definition instead.
Conda environments use the conda package manager toggle, with conda-forge as the channel.
The tool window and the Terminal both assume pip. Outside pip, follow the environment’s own package manager.
Python version and dependencies pandas requires
As of version 3.0.6, pandas needs Python 3.11 or newer (PyPI). That release was uploaded to PyPI on September 17, 2026, with wheels for CPython 3.11 through 3.15.
NumPy has a 1.26.0 minimum (pandas documentation, 3.0.6), python-dateutil has a 2.8.2 minimum, and tzdata is required only on Windows and Pyodide.
A project interpreter on 3.10 or older sits below that floor. Changing the Python version in PyCharm is the first fix, before any pip command.
Older tutorials still circulate with outdated advice. The pandas 1.5.3 documentation, the last of that series, listed Python 3.8, 3.9, 3.10 and 3.11 as supported, a range that overlaps with pandas 3.0.6 only at Python 3.11.
pip pulls in NumPy and python-dateutil on its own, so nothing needs installing first. Standalone NumPy work follows the same steps, as this guide to installing NumPy in PyCharm shows.
pytz is optional, not required. It arrives through the timezone extra, pip install "pandas[timezone]", with 2020.1 as the minimum version.
How to check that pandas installed in the right interpreter
Run import pandas as pd followed by print(pd.__version__) in a file you launch from PyCharm. If a version number prints, pandas is installed in the interpreter that run uses.
If the Run button is new to you, this guide to running code in PyCharm covers it.
There are other places to look. Settings | Python | Interpreter shows a Version and a Latest version column for every installed package (PyCharm 2026.2 Help). Open a requirements.txt and PyCharm prints the installed version next to each requirement found in the selected interpreter. In the Terminal, pip show pandas reports on the shell’s Python, which is not always the project’s.
Only the import test proves your code can reach pandas. The rest describe what one interpreter or shell contains, not which one your code runs on.
The pandas documentation also offers a test suite through pandas.test(), installed with pip install "pandas[test]". The same page warns that test failures do not necessarily mean a broken installation.
Skip it for a routine setup. The import test is enough.
Fixing the No module named pandas error in PyCharm
ModuleNotFoundError: No module named ‘pandas’ means the interpreter running your code has no pandas in its environment. The package can sit elsewhere on the machine and still trigger it.
The order matters here, so go top to bottom.
- Read the message. Pandas named in the error points to a missing pandas, while another package named in an ImportError points to a missing optional dependency.
- Open Settings | Python | Interpreter and select the interpreter your failing run uses.
- Look for pandas in its package list. If it is absent, install it there from the Python Packages tool window.
- Check that interpreter for an Invalid environment warning. If it shows, point the project at a working interpreter or rebuild the environment.
- Run the file again.
PyCharm shows Invalid environment when the specified Python binary cannot be found in the file system, or when that Python version is not supported (PyCharm 2026.2 Help). A moved or deleted environment folder fits the first case.
Clear out the broken folder before rebuilding, as this guide to deleting a venv explains.
The optional dependency case looks different. pandas raises an ImportError only when a method that needs one is called, such as read_hdf() with PyTables or to_markdown() with tabulate (pandas documentation, 3.0.6).
The import pandas line itself works in that case. Install the package the error names, or the matching extra such as pandas[hdf5].
If the Terminal reports that pip itself is missing, installing pip in PyCharm is a separate fix from everything above.
Reinstalling into the same interpreter changes nothing. Change the target, or repair the environment.
How to pin, upgrade or uninstall pandas, and install it from requirements.txt
Pin pandas with an exact specifier such as pandas==x.y.z, upgrade it with --upgrade or PyCharm’s Update button, and remove it from the tool window or with pip uninstall pandas -y. A requirements.txt file makes all of it repeatable.
Pinning a version
PyCharm’s requirements tooling accepts these specifiers:
pandas==x.y.z(strong equality) locks one releasepandas>=x.y.z(greater or equal) sets a floor and accepts anything newerpandas~=x.y.z(compatible version) allows newer releases that stay compatible
A range such as pandas>=3,<4 fits how pandas numbers releases. The project follows a loose variant of semantic versioning and reserves API-breaking changes for major releases (pandas policy page).
The policy admits one exception, since behavior-changing bug fixes can still land in minor or patch releases. A range pin narrows the risk without removing it.
If you install the timezone extra, pandas’ own pages disagree on the pytz floor. The install page lists 2020.1, while the 3.0.0 release notes list 2024.2 as the lowest tested version.
Pinning pytz>=2024.2 follows the figure the release notes call tested.
Installing from requirements.txt
- Create the file in the project root with File | New | File. requirements.txt is the recommended name.
- Type one requirement per line, for example pandas with a specifier from the list above. PyCharm completes package names as you type.
- Or, in a project that already imports pandas, open an empty requirements.txt and click Add imported packages to requirements.
Another route is pip freeze > requirements.txt in the Terminal, which writes the file from the current environment (PyCharm 2026.2 Help). It lists every installed package, not only pandas.
A root-level requirements.txt is detected automatically. A different file name, or several files, needs the Package requirements file for SDK field under Settings | Python | Tools | Integrated Tools.
Commit the file to version control so teammates install the same pins.
Upgrading and uninstalling
pip leaves an installed version alone unless --upgrade is given (pip documentation, 26.2.1). Running pip install pandas twice therefore changes nothing the second time.
- In the Terminal, run
python -m pip install --upgrade pandas - In the Python Packages tool window, use the Update button next to the package name
- In the interpreter settings, select pandas and click Upgrade
The Latest version column lists stable releases only until you click Show early releases (PyCharm 2026.2 Help).
A major jump needs more care than a patch bump. pandas 3.0.0, released January 21, 2026, removed functionality that earlier releases had deprecated (pandas 3.0.0 release notes).
Uninstalling works from the Uninstall package button in the tool window, from Uninstall on the interpreter settings toolbar, or with pip uninstall pandas -y.
FAQ About Installing pandas in PyCharm
Do you need to install wheel before pandas?
No. pip downloads a prebuilt pandas wheel from PyPI, and the pandas installation page lists no wheel step. The advice survives in older tutorials.
How do you install pandas with Excel support?
Run pip install "pandas[excel]" in the Terminal. The extra pulls in six packages, namely xlrd, xlsxwriter, openpyxl, pyxlsb, python-calamine and odfpy (pandas documentation, 3.0.6).
pandas[all] installs every optional dependency in one step.
Can you install pandas from a local file or a Git repository?
Yes. Click the Advanced Package Install button beside the search bar in the Python Packages tool window, then enter a path to a package directory, a zip or whl archive, or a Git repository (PyCharm 2026.2 Help).
The same dialog has an Editable checkbox.
Does pandas install on Windows ARM and Apple silicon?
It does. PyPI lists wheels for Windows ARM64 and for macOS ARM64 on macOS 11.0 and newer.
The pandas 3.0.0 release notes (January 21, 2026) announced the Windows ARM64 wheels.
Is PyCharm Community Edition enough to install pandas?
Going by the documentation, yes. PyCharm Community Edition is no longer a separate product: 2025.2 was its last standalone release, and JetBrains now ships a single unified PyCharm whose core features remain free (PyCharm 2026.2 Help).
The documentation labels only remote interpreters over SSH, Docker, Docker Compose and WSL as PyCharm Pro features. The Python Packages tool window and the interpreter settings are described without a Pro note.
Which pandas Version Should a Project Pin?
A new project that installs pandas in PyCharm should pin the 3.x series, since 3.x carries the current releases.
Six patch releases followed within eight months of the first 3.0 release, so a range pin collects fixes without an edit.
Existing 2.x projects need a fixed order, because 2.3 raises deprecation warnings for only a subset of the breaking changes (pandas 3.0.0 release notes).
- Reach pandas 2.3 and clear its warnings.
- Move to the 3.x series.
- Run the tests before raising the pin.
A clean 2.3 run is a prerequisite, not a guarantee, so every step ends in a test, and running pytest in PyCharm covers that gate.
The same release notes schedule further enforcement for pandas 4.0, which would reopen this advice. Position verified October 7, 2026.
- 10 Ways Shor’s Algorithm Could Impact Data Security in the Quantum Era - October 14, 2026
- VS Code Extensions That Are Actually Worth Installing - October 13, 2026
- Google Play App Signing Explained - October 11, 2026
