Updating Conda: Update Conda Itself, One Package or Every Package

Searchers for “updating conda” usually mean one of three different operations: updating conda itself, updating one package in an environment, or updating every package in an environment. Each has its own command, and mixing them up is the usual reason an update “does nothing” or changes more than intended. This guide separates them, adds a dry-run-first routine for research work, and cites the official conda documentation for every command. It was checked against the conda documentation on 2026-11-07.

Which update do you need? (quick reference)

Goal	Command	Source page
Update conda itself	conda update --name base conda (the getting-started page shows the short form conda update conda)	Managing conda; Getting started
Update one package in a named environment	conda update -n myenv scipy	conda update command reference
Update every package in an environment	conda update --all -n myenv	conda update command reference
Bring an environment in line with an environment.yml file	conda env update --file environment.yml --prune	Managing environments
Preview any of the above without changing anything	add --dry-run (-d)	conda update command reference

What conda is used for in research

Conda is a package and environment manager. Researchers use it to create separate environments, each with its own Python (or other) packages, so that one project’s analysis stack does not disturb another’s. The getting-started page shows the basic pattern: conda create --name myenvironment python numpy pandas, then conda activate myenvironment. Because an update changes the versions of the software your results were computed with, it matters for reproducibility; see the renv guide for the equivalent problem in R and mamba vs conda for a comparison of the two tools.

Install or access (official source only)

How do I install conda on macOS?

Follow the macOS instructions on the official conda installation page (docs.conda.io, “Installation”), which has separate pages for Windows, macOS and Linux. After installing, open a new terminal and run conda --version. If the terminal answers “command not found: conda”, the getting-started page says to reopen the terminal and make sure you are using the same user account that installed conda.

Step by step: how to update conda

Update conda itself (conda update conda)

1. Open a terminal and check the current version: conda --version. The documentation shows output such as conda 4.12.0; conda info and conda -V also report it.
2. Run conda update --name base conda. The getting-started page gives the equivalent conda update conda and advises running conda activate first to return to the default environment.
3. Conda reports what is available and asks whether to proceed; type y.
4. Run conda --version again and record both versions.

The same steps answer “update conda”, “conda update conda” and “how to update conda”: they are one operation under different names.

Update one package in an environment

Use conda update -n myenv scipy. The general usage on the command reference page is conda update [-h] [-n ENVIRONMENT | -p PATH] [options] [package_spec ...], where -n names an environment and -p gives its path.

How do I update all packages in a conda environment?

Run conda update --all -n myenv. The reference describes --all (also --update-all) as “Update all installed packages in the environment.” Without -n, the update applies to the currently active environment.

Update an environment from an environment.yml file

Run conda env update --file environment.yml --prune. The managing-environments page describes --prune as removing dependencies that are no longer required by the file.

How do I activate a conda environment?

Run conda activate myenv. To list the environments you have, use conda info --envs or conda env list.

A dry-run-first routine for reproducible research

1. Record conda --version and note the environment name.
2. Preview the change with conda update --all -n myenv --dry-run. The reference describes -d/--dry-run as “Only display what would have been done.”
3. Read the list of packages that would change. If a package central to your analysis would change version, decide whether you want that before analysis is finished.
4. Run the update without --dry-run (add -y to skip the confirmation prompt in scripts).
5. Record the new versions in your methods notes or project log.

The reference also lists --no-update-deps / --freeze-installed among its options, which restrict how dependencies are updated; read its description before using them.

Output and how to read it

Conda prints what it found, then the package plan, then asks for confirmation (y to proceed). With --dry-run it prints the plan and stops. After a real update, conda --version or conda list -n myenv shows the resulting versions.

Troubleshooting

• “command not found: conda”. Reopen the terminal and confirm you are in the same user account that installed conda (Getting started).
• Updated a package but the old version still appears. Check that the right environment is active (conda info --envs marks it) or pass -n explicitly.
• Update seems stuck. Conda spends time resolving a set of compatible packages before it acts. Preview with --dry-run and narrow the request to specific packages rather than --all. Consult the current conda documentation or release notes for solver behaviour in your installed version.

Free alternatives (table)

Need	Option	Where covered
Same environment management with a different tool	mamba	Mamba vs conda
Pinning package versions for R projects	renv	renv guide

FAQ

How do I update conda?

Run conda update --name base conda in a terminal and answer y when prompted (docs.conda.io, “Managing conda”).

What is conda used for?

Creating and managing environments and installing packages, so each project keeps its own software versions.

How do I update all packages in a conda env?

conda update --all -n myenv; add --dry-run first to preview.

How do I activate a conda environment?

conda activate myenv.

How do I install conda on macOS?

Use the macOS page of the official conda installation documentation, then confirm with conda --version.

For related tooling see import scipy as sp and the research tools hub.

Sources (official docs, read 2026-11-07)

• Managing conda: https://docs.conda.io/projects/conda/en/stable/user-guide/tasks/manage-conda.html
• Getting started with conda: https://docs.conda.io/projects/conda/en/stable/user-guide/getting-started.html
• conda update command reference: https://docs.conda.io/projects/conda/en/stable/commands/update.html
• Managing environments: https://docs.conda.io/projects/conda/en/stable/user-guide/tasks/manage-environments.html
• Installation: https://docs.conda.io/projects/conda/en/stable/user-guide/install/index.html