Written and maintained by CASRAI Editorial Board
Last updated
PAUP* (Phylogenetic Analysis Using PAUP) is desktop software for inferring phylogenetic trees from molecular or morphological character data using parsimony, distance and maximum-likelihood methods. This page is for researchers who need to get the program, choose the right build for their operating system and run a first analysis. (The bare word “paup” is also an English dictionary entry; this page is about the software.)
What is PAUP*?
PAUP* is a phylogenetics program written by David L. Swofford. The official site gives the name as “PAUP* (Phylogenetic Analysis Using PAUP)” and describes the site as under development (checked 2026-11-07). The original program implemented parsimony only; version 4.0, written PAUP*, added distance and likelihood methods and also includes the SVDquartets method (Wikipedia, background only). Version 4 has a graphical interface on Windows and macOS and a command-line interface; on Unix-like systems it is command-line only.
What it is used for in research
Typical tasks are building a tree from an aligned sequence matrix, comparing trees under different optimality criteria (parsimony, distance, likelihood), and saving trees for plotting elsewhere. For the concepts behind these steps see the CASRAI guide how to make a phylogenetic tree, and DNA sequencing for where alignments come from.
Install or access (official source only)
Downloads are on the PAUP* test-version page at phylosolutions.com/paup-test (read 2026-11-07). The current release listed there is 4.0a169. All versions are free to download, the expiration date has been removed from both the GUI and command-line versions, and an open-source release is planned with no date given. Choose a build by platform:
| Platform | Build and file | Notes from the download page |
|---|---|---|
| Windows | GUI installer paup4-setup.msi |
Windows 7 or later, 32- and 64-bit. The command-line version is installed with it; run paup at a command prompt. |
| macOS, graphical | PAUP_dev_icc.zip |
OS X 10.8 or later but not Catalina (10.15) or later. Use macOS 10.14 or earlier, a virtual machine, or the command-line build. On first launch, right-click and choose Open. |
| macOS, command line | paup4a168_osx.gz |
Runs on Catalina. Unzip with gunzip, then chmod a+x paup4a168_osx. |
| Linux (Intel 64-bit), command line | paup4a168_centos64.gz or paup4a169_ubuntu64.gz |
Unzip with gunzip, then chmod a+x on the file. A Fortran runtime is required (install gfortran, or R, LAPACK or Octave). |
The graphical version updates itself. Announcements are posted to the paup-announce Google group.
Run a first parsimony analysis
These steps follow the official Quick Start (paup.phylosolutions.com/tutorials/quick-start, read 2026-11-07), which uses the example file primate-mtDNA.nex.
1. Know the input format
PAUP* reads NEXUS files. The file must begin with #NEXUS, and each block is written begin ...; end;. The Quick Start describes a Data block (with dimensions ntax= nchar=, a format line and the matrix), an Assumptions block, a Paup block and a Trees block. Comments go in square brackets.
2. Load the data
- Graphical version: choose File > Execute and select
primate-mtDNA.nex. - Command line or batch:
execute primate-mtDNA.nex;
3. Search for trees
- Graphical version: choose Analysis > Heuristic Search…. The criterion (parsimony, distance or likelihood) is set under the Analysis menu.
- Command line:
hsearch;or, to start from random addition sequences,hsearch addseq=random;
4. Save the trees
- Graphical version: Trees > Save Trees to File…. Available formats include NEXUS, NEXUS without translation table, Freqpars, Newick and Hennig86.
- Command line:
savetrees file=mp.tre replace;
A short batch session
Commands taken from the Quick Start can be typed at the paup> prompt or placed in a Paup block:
log start file=practice.log replace;
set autoclose=yes warntree=no warnreset=no;
execute primate-mtDNA.nex;
hsearch addseq=random;
savetrees file=mp.tre replace;
describetrees 1 / plot=phylogram brlens=yes;
The Quick Start also shows a distance route (dset distance=hky85; then nj;) and a likelihood score command (lscore nst=2 tratio=est rates=gamma shape=estimate;). Any command’s options are listed with command-name ?.
Output and how to read it
- Tree display.
showtree;prints the current tree, anddescribetreeswithplot=phylogram brlens=yesdraws it as a phylogram with branch lengths. Usegettrees file=mp.tre;to reload saved trees. - Distances. In the graphical version, Data > Show Distance Matrix displays the pairwise distances under the current distance setting; Trees > Describe Trees… is the menu route to
describetrees. - Log file.
log start file=practice.log replace;records the session so the commands and output can be kept with the project. - Saved trees. Choose Newick when another tree viewer needs the file; NEXUS formats keep the data structure PAUP* itself reads back.
Troubleshooting
| Symptom | What the download page says |
|---|---|
| macOS graphical version will not run on Catalina or later | It is not supported from macOS 10.15. Use macOS 10.14 or earlier, a virtual machine, or the command-line build. |
| macOS blocks the first launch | Right-click the application and choose Open. |
| macOS launch error about a missing font | Quit the fontd process in Activity Monitor. |
| Mac or Linux command-line file will not execute | Run gunzip on the download, then chmod a+x on the file. |
| Linux build fails to start | A Fortran runtime is needed: install gfortran, or R, LAPACK or Octave. |
| Manual does not match the program | The draft manual PDF (phylosolutions.com/paup-documentation/paupmanual.pdf) is described by its authors as incomplete and stale. Use the Quick Start and the built-in command-name ? help for current syntax. |
Alternative phylogenetics software
Each row states only what the project’s own page says (read 2026-11-07).
| Program | What the project says it does | Interface and platforms |
|---|---|---|
| IQ-TREE | “Efficient software for phylogenomic inference”: maximum-likelihood tree search, ModelFinder model selection, UFBoot ultrafast bootstrap support. | Release 3.1.4 for Windows, macOS and Linux (64-bit). |
| MEGA | Software suite for DNA and protein sequence analysis, with maximum likelihood, distance, maximum parsimony and Bayesian methods, plus alignment and tree-viewing tools. | Graphical and command-line versions; Windows, macOS and Linux. |
| RAxML-NG | Maximum-likelihood tree inference, successor to RAxML, with model selection and bootstrap support. Licensed AGPL-3.0. | Command line; pre-built Linux and macOS binaries, Windows via WSL. |
For other tools in the cluster see free statistical software.
FAQ
What is a PAUP?
In research, PAUP* is the phylogenetics program “Phylogenetic Analysis Using PAUP” by David L. Swofford. It builds trees from character data using parsimony, distance and likelihood methods.
Where do I download PAUP*?
From phylosolutions.com/paup-test. As of 2026-11-07 the current release is 4.0a169 and all versions are free to download.
Is there a PAUP* manual?
A draft manual PDF exists but is described as incomplete and stale. The official Quick Start and in-program help are the current references.
Does PAUP* have a GUI?
Yes on Windows (the installer includes the command-line version) and on macOS up to 10.14. Linux has command-line builds only.
Is there a PAUP* app for newer Macs?
The graphical macOS build does not run on Catalina (10.15) or later; the command-line macOS build does.
Sources
- PAUP* official site, read 2026-11-07
- PAUP* test version downloads, read 2026-11-07
- PAUP* Quick Start, read 2026-11-07
- Wikipedia: PAUP (background only), read 2026-11-07
- IQ-TREE, MEGA, RAxML-NG, read 2026-11-07







