> For the complete documentation index, see [llms.txt](https://developer.harness.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.harness.io/ai-test-automation/use-ai-test-automation/test-authoring/create-tests/conditionals.md).
# Conditionals
Reference for conditional branching and guaranteed execution blocks in Harness AI Test Automation, including IF, ELSE, and FINALLY statements
## Overview
Conditionals enable you to create dynamic test flows with different execution paths based on specified conditions. Harness AI Test Automation provides two types of control flow mechanisms:
**Conditional Branching**: Use IF, ELSE IF, ELSE, and END IF statements to build branching logic that executes different steps based on runtime conditions.
**Guaranteed Execution**: Use FINALLY blocks to ensure critical cleanup or verification steps always run at the end of a test, regardless of whether the test passes or fails.
## Conditional statements
### IF
Creates conditional branches in test execution based on specified conditions.
| Feature | Description |
| ------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Description** | Evaluates a condition that must return TRUE or FALSE. If the condition evaluates to true, the subsequent steps within the conditional block are executed. |
| **Parameter option - Question** |
• User question: Prompts the user with a question during test execution • Text exists: Checks if specific text is present on the page • Text does not exist: Verifies that specific text is not present on the page • Use custom script: Accepts custom JavaScript that evaluates to TRUE or FALSE
|
| **Advanced options** |
• Use page HTML as context: When enabled, evaluates the condition against the page's HTML source • Disable AI code generation for this assertion: When enabled, causes Copilot to rely solely on visual inspection.
|
| **Return value** | TRUE or FALSE |
{% hint style="info" %}
**USAGE**\
Every IF block must be closed with an END IF statement. Tests with unclosed IF blocks cannot be saved.
{% endhint %}
### ELSE
Creates an alternative execution path.
| Feature | Description |
| --------------------- | ------------------------------------------------------------------------------------------------------------------- |
| **Description** | Used to create an alternative path for test execution when the preceding IF or ELSE IF conditions evaluate to FALSE |
| **Parameter options** | None |
| **Advanced options** | None |
| **Return value** | None |
{% hint style="info" %}
**USAGE**
ELSE must be preceded by an IF or ELSE IF statement within the same conditional block.
{% endhint %}
### ELSE IF
Creates an alternative execution path.
| Feature | Description |
| ------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Description** | Evaluates an additional condition when the preceding IF or ELSE IF conditions evaluate to FALSE. If this condition returns true, the subsequent steps within this block are executed. |
| **Parameter option - Question** |
• User question: Prompts the user with a question during test execution • Text exists: Checks if specific text is present on the page • Text does not exist: Verifies that specific text is not present on the page • Use custom script: Accepts custom JavaScript that evaluates to TRUE or FALSE
|
| **Advanced options** |
• Use page HTML as context: When enabled, evaluates the condition against the page's HTML source • Disable AI code generation for this assertion: When enabled, causes Copilot to rely solely on visual inspection.
|
| **Return value** | TRUE or FALSE |
{% hint style="info" %}
**USAGE**\
ELSE IF must be preceded by an IF or another ELSE IF statement within the same conditional block.
{% endhint %}
### END IF
Marks the end of a conditional block, completing the IF/ELSE IF/ELSE structure.
| Feature | Description |
| --------------------- | ------------------------------------------------------------------------------------- |
| **Description** | Used to end a conditional block, marking the conclusion of IF/ELSE IF/ELSE statements |
| **Parameter options** | None |
| **Advanced options** | None |
| **Return value** | None |
{% hint style="info" %}
**USAGE**\
END IF must be preceded by an IF, ELSE IF, or ELSE statement to close the conditional block.
{% endhint %}
### FINALLY
Ensures critical cleanup or verification steps are always executed, regardless of test outcome.
| Feature | Description |
| --------------------- | -------------------------------------------------------------------------------------------------------- |
| **Description** | Defines a block of steps that will always execute at the end of a test, whether the test passes or fails |
| **Parameter options** | None |
| **Advanced options** | None |
| **Return value** | None |
{% hint style="info" %}
**INFO**
* **Scope**: Each test can have at most one FINALLY block.
* **Can contain**: Any test steps, including IF/ELSE conditionals.
* **Restrictions**: Cannot be nested inside conditional blocks or used within [tasks](/ai-test-automation/use-ai-test-automation/test-authoring/create-tests/tasks.md).
{% endhint %}
> Use FINALLY blocks to ensure your test environment is always left in a clean state, even when unexpected failures occur. This helps maintain test reliability and prevents side effects that could impact subsequent test runs.
{% @harness-feedback/feedback %}
