First, install behave.

Now make a directory called “features”. In that directory create a file called “tutorial.feature” containing:

Feature: showing off behave

  Scenario: run a simple test
     Given we have behave installed
      When we implement a test
      Then behave will test it for us!

Make a new directory called “features/steps”. In that directory create a file called “tutorial.py” containing:

from behave import *

@given('we have behave installed')
def step_impl(context):

@when('we implement a test')
def step_impl(context):
    assert True is not False

@then('behave will test it for us!')
def step_impl(context):
    assert context.failed is False

Run behave:

% behave
Feature: showing off behave # features/tutorial.feature:1

  Scenario: run a simple test        # features/tutorial.feature:3
    Given we have behave installed   # features/steps/tutorial.py:3
    When we implement a test         # features/steps/tutorial.py:7
    Then behave will test it for us! # features/steps/tutorial.py:11

1 feature passed, 0 failed, 0 skipped
1 scenario passed, 0 failed, 0 skipped
3 steps passed, 0 failed, 0 skipped, 0 undefined

Now, continue reading to learn how to make the most of behave.


behave operates on directories containing:

  1. feature files written by your Business Analyst / Sponsor / whoever with your behaviour scenarios in it, and
  2. a “steps” directory with Python step implementations for the scenarios.

You may optionally include some environmental controls (code to run before and after steps, scenarios, features or the whole shooting match).

The minimum requirement for a features directory is:


A more complex directory might look like:


If you’re having trouble setting things up and want to see what behave is doing in attempting to find your features use the “-v” (verbose) command-line switch.

Feature Files

A feature file has a natural language format describing a feature or part of a feature with representative examples of expected outcomes. They’re plain-text (encoded in UTF-8) and look something like:

Feature: Fight or flight
  In order to increase the ninja survival rate,
  As a ninja commander
  I want my ninjas to decide whether to take on an
  opponent based on their skill levels

  Scenario: Weaker opponent
    Given the ninja has a third level black-belt
     When attacked by a samurai
     Then the ninja should engage the opponent

  Scenario: Stronger opponent
    Given the ninja has a third level black-belt
     When attacked by Chuck Norris
     Then the ninja should run for his life

The “Given”, “When” and “Then” parts of this prose form the actual steps that will be taken by behave in testing your system. These map to Python step implementations. As a general guide:

Given we put the system in a known state before the user (or external system) starts interacting with the system (in the When steps). Avoid talking about user interaction in givens.

When we take key actions the user (or external system) performs. This is the interaction with your system which should (or perhaps should not) cause some state to change.

Then we observe outcomes.

You may also include “And” or “But” as a step - these are renamed by behave to take the name of their preceding step, so:

Scenario: Stronger opponent
  Given the ninja has a third level black-belt
   When attacked by Chuck Norris
   Then the ninja should run for his life
    And fall off a cliff

In this case behave will look for a step definition for "Then fall off a cliff".

Scenario Outlines

Sometimes a scenario should be run with a number of variables giving a set of known states, actions to take and expected outcomes, all using the same basic actions. You may use a Scenario Outline to achieve this:

Scenario Outline: Blenders
   Given I put <thing> in a blender,
    when I switch the blender on
    then it should transform into <other thing>

 Examples: Amphibians
   | thing         | other thing |
   | Red Tree Frog | mush        |

 Examples: Consumer Electronics
   | thing         | other thing |
   | iPhone        | toxic waste |
   | Galaxy Nexus  | toxic waste |

behave will run the scenario once for each (non-heading) line appearing in the example data tables.

Step Data

Sometimes it’s useful to associate a table of data with your step.

Any text block following a step wrapped in """ lines will be associated with the step. For example:

Scenario: some scenario
  Given a sample text loaded into the frobulator
     Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do
     eiusmod tempor incididunt ut labore et dolore magna aliqua.
 When we activate the frobulator
 Then we will find it similar to English

The text is available to the Python step code as the “.text” attribute in the Context variable passed into each step function.

You may also associate a table of data with a step by simply entering it, indented, following the step. This can be useful for loading specific required data into a model.

Scenario: some scenario
  Given a set of specific users
     | name      | department  |
     | Barry     | Beer Cans   |
     | Pudey     | Silly Walks |
     | Two-Lumps | Silly Walks |

 When we count the number of people in each department
 Then we will find two people in "Silly Walks"
  But we will find one person in "Beer Cans"

The table is available to the Python step code as the “.table” attribute in the Context variable passed into each step function. The table for the example above could be accessed like so:

@given('a set of specific users')
def step_impl(context):
    for row in context.table:
        model.add_user(name=row['name'], department=row['department'])

There’s a variety of ways to access the table data - see the Table API documentation for the full details.

Python Step Implementations

Steps used in the scenarios are implemented in Python files in the “steps” directory. You can call these whatever you like as long as they use the python *.py file extension. You don’t need to tell behave which ones to use - it’ll use all of them.

The full detail of the Python side of behave is in the API documentation.

Steps are identified using decorators which match the predicate from the feature file: given, when, then and step (variants with Title case are also available if that’s your preference.) The decorator accepts a string containing the rest of the phrase used in the scenario step it belongs to.

Given a Scenario:

Scenario: Search for an account
   Given I search for a valid account
    Then I will see the account details

Step code implementing the two steps here might look like (using selenium webdriver and some other helpers):

@given('I search for a valid account')
def step_impl(context):
    form = get_element(context.browser, tag='form')
    get_element(form, name="msisdn").send_keys('61415551234')

@then('I will see the account details')
def step_impl(context):
    elements = find_elements(context.browser, id='no-account')
    eq_(elements, [], 'account not found')
    h = get_element(context.browser, id='account-head')
    ok_(h.text.startswith("Account 61415551234"),
        'Heading %r has wrong text' % h.text)

The step decorator matches the step to any step type, “given”, “when” or “then”. The “and” and “but” step types are renamed internally to take the preceding step’s keyword (so an “and” following a “given” will become a “given” internally and use a given decorated step).


Step function names do not need to have a unique symbol name, because the text matching selects the step function from the step registry before it is called as anonymous function. Hence, when behave prints out the missing step implementations in a test run, it uses “step_impl” for all functions by default.

If you find you’d like your step implementation to invoke another step you may do so with the Context method execute_steps().

This function allows you to, for example:

@when('I do the same thing as before')
def step_impl(context):
        when I press the big red button
         and I duck

This will cause the “when I do the same thing as before” step to execute the other two steps as though they had also appeared in the scenario file.

Step Parameters

You may find that your feature steps sometimes include very common phrases with only some variation. For example:

Scenario: look up a book
  Given I search for a valid book
   Then the result page will include "success"

Scenario: look up an invalid book
  Given I search for a invalid book
   Then the result page will include "failure"

You may define a single Python step that handles both of those Then clauses (with a Given step that puts some text into context.response):

@then('the result page will include "{text}"')
def step_impl(context, text):
    if text not in context.response:
        fail('%r not in %r' % (text, context.response))

There are several parsers available in behave (by default):

parse (the default, based on: parse)

Provides a simple parser that replaces regular expressions for step parameters with a readable syntax like {param:Type}. The syntax is inspired by the Python builtin string.format() function. Step parameters must use the named fields syntax of parse in step definitions. The named fields are extracted, optionally type converted and then used as step function arguments.

Supports type conversions by using type converters (see register_type()).

cfparse (extends: parse, requires: parse_type)

Provides an extended parser with “Cardinality Field” (CF) support. Automatically creates missing type converters for related cardinality as long as a type converter for cardinality=1 is provided. Supports parse expressions like:

  • {values:Type+} (cardinality=1..N, many)
  • {values:Type*} (cardinality=0..N, many0)
  • {value:Type?} (cardinality=0..1, optional).

Supports type conversions (as above).


This uses full regular expressions to parse the clause text. You will need to use named groups “(?P<name>…)” to define the variables pulled from the text and passed to your step() function.

Type conversion is not supported. A step function writer may implement type conversion inside the step function (implementation).

To specify which parser to use invoke use_step_matcher() with the name of the matcher to use. You may change matcher to suit specific step functions - the last call to use_step_matcher before a step function declaration will be the one it uses.


The function step_matcher() is becoming deprecated. Use use_step_matcher() instead.


You’ll have noticed the “context” variable that’s passed around. It’s a clever place where you and behave can store information to share around. It runs at three levels, automatically managed by behave.

When behave launches into a new feature or scenario it adds a new layer to the context, allowing the new activity level to add new values, or overwrite ones previously defined, for the duration of that activity. These can be thought of as scopes.

You can define values in your environmental controls file which may be set at the feature level and then overridden for some scenarios. Changes made at the scenario level won’t permanently affect the value set at the feature level.

You may also use it to share values between steps. For example, in some steps you define you might have:

@given('I request a new widget for an account via SOAP')
def step_impl(context):
    client = Client("")
    context.response = client.Allocate(customer_first='Firstname',
        customer_last='Lastname', colour='red')

@then('I should receive an OK SOAP response')
def step_impl(context):
    eq_(context.response['ok'], 1)

There’s also some values added to the context by behave itself:

This holds any table data associated with a step.
This holds any multi-line text associated with a step.
This is set at the root of the context when any step fails. It is sometimes useful to use this combined with the --stop command-line option to prevent some mis-behaving resource from being cleaned up in an after_feature() or similar (for example, a web browser being driven by Selenium.)

The context variable in all cases is an instance of behave.runner.Context.

Environmental Controls

The environment.py module may define code to run before and after certain events during your testing:

before_step(context, step), after_step(context, step)
These run before and after every step.
before_scenario(context, scenario), after_scenario(context, scenario)
These run before and after each scenario is run.
before_feature(context, feature), after_feature(context, feature)
These run before and after each feature file is exercised.
before_tag(context, tag), after_tag(context, tag)
These run before and after a section tagged with the given name. They are invoked for each tag encountered in the order they’re found in the feature file. See controlling things with tags.
before_all(context), after_all(context)
These run before and after the whole shooting match.

The feature, scenario and step objects represent the information parsed from the feature file. They have a number of attributes:

“Feature”, “Scenario”, “Given”, etc.
The name of the step (the text after the keyword.)
A list of the tags attached to the section or step. See controlling things with tags.
filename and line
The file name (or “<string>”) and line number of the statement.

A common use-case for environmental controls might be to set up a web server and browser to run all your tests in. For example:

# -- FILE: features/environment.py
from behave import fixture, use_fixture
from behave4my_project.fixtures import wsgi_server
from selenium import webdriver

def selenium_browser_chrome(context):
    # -- HINT: @behave.fixture is similar to @contextlib.contextmanager
    context.browser = webdriver.Chrome()
    yield context.browser

def before_all(context):
    use_fixture(wsgi_server, context, port=8000)
    use_fixture(selenium_browser_chrome, context)
    # -- HINT: CLEANUP-FIXTURE is performed after after_all() hook is called.

def before_feature(context, feature):
# -- FILE: behave4my_project/fixtures.py
# ALTERNATIVE: Place fixture in "features/environment.py" (but reuse is harder)
from behave import fixture
import threading
from wsgiref import simple_server
from my_application import model
from my_application import web_app

def wsgi_server(context, port=8000):
    context.server = simple_server.WSGIServer(('', port))
    context.thread = threading.Thread(target=context.server.serve_forever)
    yield context.server

Of course, if you wish, you could have a new browser for each feature, or to retain the database state between features or even initialise the database for each scenario.

Controlling Things With Tags

You may also “tag” parts of your feature file. At the simplest level this allows behave to selectively check parts of your feature set.

Given a feature file with:

Feature: Fight or flight
  In order to increase the ninja survival rate,
  As a ninja commander
  I want my ninjas to decide whether to take on an
  opponent based on their skill levels

  Scenario: Weaker opponent
    Given the ninja has a third level black-belt
    When attacked by a samurai
    Then the ninja should engage the opponent

  Scenario: Stronger opponent
    Given the ninja has a third level black-belt
    When attacked by Chuck Norris
    Then the ninja should run for his life

then running behave --tags=slow will run just the scenarios tagged @slow. If you wish to check everything except the slow ones then you may run behave --tags=-slow.

Another common use-case is to tag a scenario you’re working on with @wip and then behave --tags=wip to just test that one case.

Tag selection on the command-line may be combined:

  • --tags=wip,slow
    This will select all the cases tagged either “wip” or “slow”.
  • --tags=wip --tags=slow
    This will select all the cases tagged both “wip” and “slow”.

If a feature or scenario is tagged and then skipped because of a command-line control then the before_ and after_ environment functions will not be called for that feature or scenario. Note that behave has additional support specifically for testing works in progress.

The tags attached to a feature and scenario are available in the environment functions via the “feature” or “scenario” object passed to them. On those objects there is an attribute called “tags” which is a list of the tag names attached, in the order they’re found in the features file.

There are also environmental controls specific to tags, so in the above example behave will attempt to invoke an environment.py function before_tag and after_tag before and after the Scenario tagged @slow, passing in the name “slow”. If multiple tags are present then the functions will be called multiple times with each tag in the order they’re defined in the feature file.

Re-visiting the example from above; if only some of the features required a browser and web server then you could tag them @browser:

# -- FILE: features/environment.py
# HINT: Reusing some code parts from above.

def before_feature(context, feature):
    if 'browser' in feature.tags:
        use_fixture(wsgi_server, context)
        use_fixture(selenium_browser_chrome, context)

Works In Progress

behave supports the concept of a highly-unstable “work in progress” scenario that you’re actively developing. This scenario may produce strange logging, or odd output to stdout or just plain interact in unexpected ways with behave’s scenario runner.

To make testing such scenarios simpler we’ve implemented a “-w” command-line flag. This flag:

  1. turns off stdout capture

  2. turns off logging capture; you will still need to configure your own logging handlers - we recommend a before_all() with:

    if not context.config.log_capture:
  3. turns off pretty output - no ANSI escape sequences to confuse your scenario’s output

  4. only runs scenarios tagged with “@wip”

  5. stops at the first error


Fixtures simplify the setup/cleanup tasks that are often needed during test execution.

# -- FILE: behave4my_project/fixtures.py  (or in: features/environment.py)
from behave import fixture
from somewhere.browser.firefox import FirefoxBrowser

# -- FIXTURE: Use generator-function
def browser_firefox(context, timeout=30, **kwargs):
    context.browser = FirefoxBrowser(timeout, **kwargs)
    yield context.browser

See Fixtures for more information.

Debug-on-Error (in Case of Step Failures)

A “debug on error/failure” functionality can easily be provided, by using the after_step() hook. The debugger is started when a step fails.

It is in general a good idea to enable this functionality only when needed (in interactive mode). The functionality is enabled (in this example) by using the user-specific configuration data. A user can:

  • provide a userdata define on command-line
  • store a value in the “behave.userdata” section of behave’s configuration file
# -- FILE: features/environment.py
# USE: behave -D BEHAVE_DEBUG_ON_ERROR         (to enable  debug-on-error)
# USE: behave -D BEHAVE_DEBUG_ON_ERROR=yes     (to enable  debug-on-error)
# USE: behave -D BEHAVE_DEBUG_ON_ERROR=no      (to disable debug-on-error)


def setup_debug_on_error(userdata):

def before_all(context):

def after_step(context, step):
    if BEHAVE_DEBUG_ON_ERROR and step.status == "failed":
        # -- ENTER DEBUGGER: Zoom in on failure location.
        # NOTE: Use IPython debugger, same for pdb (basic python debugger).
        import ipdb