Getting started

Generally, you will want to install pymssql with:

pip install pymssql

FreeTDS is required. On some platforms, we provide a pre-compiled FreeTDS to make installing easier, but you may want to install FreeTDS before doing pip install pymssql if you run into problems or need features or bug fixes in a newer version of FreeTDS. You can build FreeTDS from source if you want the latest. If you’re okay with the latest version that your package manager provides, then you can use your package manager of choice to install FreeTDS. E.g.:

  • Ubuntu/Debian:

    sudo apt-get install freetds-dev
  • Mac OS X with Homebrew:

    brew install freetds



Another possible way to get started quickly with pymssql is to use a Docker image.

See the Docker docs for installation instructions for a number of platforms; you can try this link:

There is a pymssql docker image on the Docker Registry at:

It is a Docker image with:

  • Ubuntu 14.04 LTS (trusty)
  • Python 2.7.6
  • pymssql
  • FreeTDS 0.91
  • SQLAlchemy 0.9.8
  • Alembic 0.7.4
  • Pandas 0.15.2
  • Numpy 1.9.1
  • IPython 2.3.1

To try it, first download the image (this requires Internet access and could take a while):

docker pull pymssql/pymssql

Then run a Docker container using the image with:

docker run -it --rm pymssql/pymssql

By default, if no command is specified, an IPython shell is invoked. You can override the command if you wish – e.g.:

docker run -it --rm pymssql/pymssql bin/bash

Here’s how using the Docker container looks in practice:

$ docker pull pymssql/pymssql
$ docker run -it --rm pymssql/pymssql
Python 2.7.6 (default, Mar 22 2014, 22:59:56)
Type "copyright", "credits" or "license" for more information.

IPython 2.1.0 -- An enhanced Interactive Python.
?         -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help      -> Python's own help system.
object?   -> Details about 'object', use 'object??' for extra details.

In [1]: import pymssql; pymssql.__version__
Out[1]: u'2.1.1'

In [2]: import sqlalchemy; sqlalchemy.__version__
Out[2]: '0.9.7'

In [3]: import pandas; pandas.__version__
Out[3]: '0.14.1'



The pymssql package consists of two modules:

  • pymssql – use it if you care about DB-API compliance, or if you are accustomed to DB-API syntax,
  • _mssql – use it if you care about performance and ease of use (_mssql module is easier to use than pymssql).

And, as of version 2.1.x it uses the services of the db-lib component of FreeTDS. See the relevant FreeTDS documentation for additional details.


Remember to install FreeTDS first.


pip install pymssql

will install pymssql from PyPI. This PyPI page contains:

  • source distribution (.tar.gz)
  • wheels (.whl) for Windows

Project Discussion

Discussions and support take place on pymssql mailing list here:, you can participate via web, e-mail or read-only subscribing to the mailing list feeds.

This is the best place to get help, please feel free to drop by and ask a question.

Project Status

Current release: 2.x is the branch under current development. It is a complete rewrite using Cython and the latest FreeTDS libraries (which remove many of the limitations of previous versions of FreeTDS).

Legacy release: 1.0.3 is the legacy version and is no longer under active development.


This documentation is for pymssql 2.x.

The document set you are reading describes exclusively the code base of pymssql 2.x and newer. All description of functionality, workarounds, limitations, dependencies, etc. of older revisions has been removed.

If you need help for building/using pymssql 1.x please refer to the old Google Code documentation Wiki.

Current Development

Official development repositories and issue trackers have been moved to GitHub at

We would be happy to have:

  • A couple more developers
  • Help from the community with maintenance of this documentation.

If interested, please connect with us on the mailing list.