Sample Test User Data: Names, Emails, Phones, and Addresses by Country
A QA-ready reference of safe sample test user data — names, emails, phone numbers, and addresses by country — that will not collide with real users or violate carrier rules.
· 8 min read
Sooner or later every QA engineer needs a sample test user that looks real enough to pass validation but is guaranteed not to belong to an actual person. This reference collects the safe ranges and reserved values for names, emails, phone numbers, and addresses across the most common locales.
Test email addresses
RFC 2606 reserves four top-level domains for testing: .test, .example, .invalid, and .localhost. Anything @example.com, @example.org, or @example.net is also reserved and will never be assigned to a real user. For inbox testing, services like Mailosaur, Mailtrap, and Ethereal generate disposable addresses you can read programmatically.
Test phone numbers — North America
The North American Numbering Plan reserves the 555-0100 through 555-0199 range in every area code for fictitious use. So +1 (212) 555-0123 is safe to use in any test. For SMS testing, Twilio provides magic numbers like +1 500 555 0006 (always passes verification) in their sandbox.
Test phone numbers — UK and EU
Ofcom in the UK reserves 07700 900000 to 07700 900999 for drama and testing — these will never reach a real handset. For Germany, the 0152 28817386 range and similar 0152/0157 ranges are unassigned. France reserves the 06 39 98 xx xx and 09 80 xx xx xx ranges for fiction. Always check the regulator's current list before going live.
Test addresses
Use addresses of public landmarks (10 Downing Street, 1600 Pennsylvania Avenue) only for visual mock-ups, never for KYC or address-validation tests. For validation tests, payment providers like Stripe and Adyen publish test addresses that always pass AVS — Stripe's '1 Embarcadero Center, San Francisco, CA 94111' is a safe default.
Test names
Names are the easiest to fake but also the easiest to get wrong. Use Faker locales (faker.name.firstName('ja') for Japanese, 'de' for German) to get culturally plausible names with the right diacritics. For edge cases, keep a small library of pathological names: O'Brien, Müller, 山田, von der Leyen, and a single-word mononym like Madonna. They flush out validation bugs faster than any random generator.
Make these one keystroke away
Looking up safe ranges every time you write a test wastes the whole point of having them. Store the sets your team actually uses — five names per locale, two phones per country, the AVS-passing addresses for each provider — in a shared TestAssets list and paste them into any field with a shortcut. The reference stays canonical; the friction disappears.