End-to-End Testing

What is End-to-End Testing?

End-to-end (E2E) testing verifies complete user journeys, such as signing up or checking out, through the real interface, services, databases and integrations together. These tests are slow and brittle, so teams keep few.

Updated November 2026 · 3 min read


End-to-end (E2E) testing checks that a complete user workflow works from start to finish across every part of the system involved: the user interface, back-end services, databases, third-party integrations and the network between them. Where a unit test asks "does this function return the right value?", an end-to-end test asks "can a customer actually sign up, add an item to the cart and pay?"

E2E tests catch problems that only appear when components run together, such as a front end sending a field the API no longer accepts, a background job that fails silently, or a payment provider returning an unexpected response. They give the highest confidence that the product works for real users, but they are also the slowest, most expensive and most fragile tests to maintain, so teams use them selectively.

Where E2E tests fit in the test pyramid

The test pyramid, popularized by Mike Cohn in Succeeding with Agile (2009), recommends many fast, cheap tests at the bottom and few slow, expensive ones at the top.

LevelWhat it checksSpeedTypical share
Unit testsA single function or class in isolationMillisecondsMost tests
Integration testsTwo or more components working together, such as a service and its databaseSecondsA moderate number
End-to-end testsA full user journey through the deployed systemSeconds to minutes eachA small number of critical paths

Types of end-to-end testing

  • Horizontal E2E testing. Follows a user journey across several applications or services, such as an order flowing from a web store to the warehouse system to the shipping provider.
  • Vertical E2E testing. Tests every layer of a single feature, from the UI through the API and business logic to the database.
  • Manual vs. automated. Manual E2E checks are common before major releases and during user acceptance testing. Automated E2E suites run in the CI pipeline using browser-automation tools such as Selenium, Cypress or Playwright.

How to build an end-to-end test suite

  1. List the critical journeys. Choose the workflows where failure would hurt most: sign-up, login, checkout, core data creation.
  2. Write each journey as user steps. Describe what the user does and what they should see, with expected results at each step.
  3. Prepare test data and environments. Use a production-like environment and data that can be reset between runs.
  4. Automate the stable journeys. Automate flows that change rarely; keep checks of fast-changing screens manual or lower in the pyramid.
  5. Run them in the pipeline. Run E2E tests on every merge to main or before every deploy, with clear failure reports and screenshots.
  6. Fix or remove flaky tests quickly. A test that fails randomly trains people to ignore failures.

A hypothetical example: Checkout

Consider a hypothetical online retailer with 1,200 unit tests, 150 integration tests and 25 end-to-end tests. The unit tests run in three minutes. The E2E tests average 48 seconds each, so running them one after another takes 25 × 48 = 1,200 seconds, or 20 minutes. Split across four parallel workers, they finish in about five minutes, which is fast enough to run before every deploy.

One E2E test covers guest checkout: open the home page, search for a product, add it to the cart, enter a shipping address, pay with a test card and confirm that an order confirmation appears and an order record exists. During one release, every unit test passes but this E2E test fails: a change to the address service renamed a field, and the payment step no longer receives the postal code. The bug is caught before deployment instead of by customers.

Common pitfalls

  • Testing everything end to end. Hundreds of E2E tests create a slow, brittle suite. Push detailed checks down to unit and integration tests.
  • Tolerating flaky tests. Intermittent failures caused by timing or shared data erode trust in the whole suite.
  • Unrealistic environments. Tests that pass against mocks may still fail against real integrations. Keep at least some E2E runs in a production-like setup.
  • No ownership. E2E tests span teams, so they are easy to orphan. Assign each journey an owner.

End-to-end testing is one part of an automated testing strategy and sits naturally inside continuous integration pipelines. It complements functional testing and the broader discipline of quality assurance.

Frequently asked questions

What is the difference between end-to-end testing and integration testing?

Integration tests check that two or more components work together, such as a service and its database, and usually run in seconds. End-to-end tests follow a full user journey through the deployed system, including the interface, services, databases and third-party integrations, and take seconds to minutes each. Teams keep a moderate number of integration tests and only a few E2E tests.

Which tools are used for automated end-to-end testing?

Automated end-to-end suites commonly use browser-automation tools such as Selenium, Cypress or Playwright, running in the CI pipeline on every merge to main or before every deploy. Manual E2E checks remain common before major releases and during user acceptance testing, and for screens that change too often for automated tests to be worth maintaining.

How many end-to-end tests should a team have?

Keep end-to-end tests to a small number covering the journeys where failure would hurt most, such as sign-up, login, checkout and core data creation. The test pyramid, popularized by Mike Cohn, puts most tests at the unit testing level, a moderate number at the integration level and few E2E tests on top, because they are slow, costly and fragile.