US Business Banking for Non-Residents: How Accounts Actually Get Opened

Most non-resident LLCs do not fail at formation. They fail at the bank.

Forming a Wyoming LLC is a filing. Getting a US institution to hold your money is a risk decision made by a human or a model that has never met you, based entirely on what your paperwork says about you. That asymmetry is the reason this page exists.

I have run my own US LLC since 2019 and built and maintained more than 120 for founders outside the United States. The banking step is where the checkout-model platforms quietly disappear: they hand you a link to a fintech marketplace and call it “banking support”. What follows is what actually determines the outcome.

The honest version: nobody can guarantee you a US bank account. The decision is always the bank’s. What can be engineered is everything that reaches their desk before they decide.

The two tiers of US banking

Founders talk about “a US bank account” as if it were one product. It is two, with different rules, different timelines and different failure modes. The table above summarises it; here is what it means in practice.

Fintechs (Mercury, Relay, Wise) are built for remote onboarding. They will open in 48 to 72 hours once your LLC and EIN are in place and your application is coherent. The trade-off is that their risk models are automated and unforgiving: accounts get frozen or closed with little warning when activity stops matching the stated business.

Traditional banks (Chase, Bank of America, Citi) are the opposite. Slow, documentation-heavy, usually requiring an in-branch appointment and sometimes an ITIN. But once you are in, you are in: the relationship is durable, it unlocks credit, and it carries weight with enterprise clients and payment processors.

The right answer for most founders is sequential, not either-or: open with a fintech to become operational, then add a traditional relationship when there is a reason to.

What the bank is actually checking

Every institution runs a version of the same four questions. Nothing here is secret; what is rare is having all four answered cleanly before you apply.

• Is this entity real and in good standing? Articles of Organization, an EIN confirmation letter (CP 575), a registered agent that exists, state filings up to date.
• Who controls it, and can that be verified? Passport, proof of address, ownership structure with no unexplained layers.
• Does the business story hold together? Stated activity, expected volume, client geography and the operating agreement all describing the same company. Incoherence here is the single most common cause of decline.
• Does the risk profile fit their book? Some sectors, some countries and some volume patterns are outside a given institution’s appetite regardless of how good your file is. That is not a rejection of you; it is a mismatch of bank.

The application, step by step

Skip past the marketing pages and the sequence is the same everywhere, fintech or traditional. Getting it in the wrong order is how founders manufacture their own decline.

1. Form the LLC and receive the Articles of Organization. Nothing downstream starts before this exists.
2. Obtain the EIN and wait for it to settle. The IRS issues the number quickly in some cases and takes weeks in others when there is no SSN on the application. A freshly minted EIN, days old, can add friction, because the institution cannot yet cross-check your entity against external data sources. It is not disqualifying; it is a reason to build in a few days of runway rather than applying the hour the confirmation letter arrives.
3. Assemble the file before opening any application form. Passport, Articles of Organization, EIN confirmation (CP 575), an operating agreement that actually describes the business, a coherent business address, and the business description itself.
4. Apply once, cleanly. One institution, one attempt, from a connection consistent with your stated residence. Repeated attempts in quick succession, or applications filed through an aggressive VPN, read as evasive rather than cautious.
5. Respond fast if the institution asks follow-up questions. Silence during review reads worse than an imperfect answer.

What each type of institution actually asks for, concretely:

• Fintechs (Mercury, Relay) want the EIN confirmation, the passport of the beneficial owner, the Articles of Organization, and a business description that a compliance analyst, not a domain expert, can understand in one read. They verify identity electronically and cross-reference the entity against IRS records; there is no human interview in most cases.
• Wise Business wants broadly the same document set, plus more detail on where funds originate and where they are going if your flows are multi-currency.
• Traditional banks want everything a fintech wants, plus in most cases an in-person appointment, sometimes an SSN or ITIN, and often a US mailing address or a banker relationship that vouches for the file before you walk in.

Writing a business description that clears review

This is the single detail that saves or sinks more applications than anything else on this page, and almost nobody spends real time on it. The business description is what a compliance analyst reads to decide whether your company is coherent. It has to answer three questions in one or two sentences: what do you sell, to whom, and how do you get paid.

• Weak: “Consulting and services.”
• Strong: “Web design services for small businesses in the United States, billed per project via Stripe and ACH transfer.”

Name the sector, name the client type (ideally with a US presence if that is genuinely your market), and name the payment method. Avoid language that triggers automated flags: certain sectors (crypto, gambling, adult, high-risk MSB activity) draw manual review regardless of how the rest of the file looks, so do not reach for vague, evasive phrasing if that is not your business; and do not borrow language from a different industry because it sounds more impressive. The description has to be true and it has to be verifiable against your website. When the business description, the operating agreement, and the actual transaction pattern all tell the same story, review moves fast. When they diverge even slightly, that divergence is the first thing an analyst notices.

Why applications get declined

In my experience the declines cluster into a handful of patterns, and almost all of them are preventable:

• A template operating agreement with the founder’s name dropped into it, describing a company that does not match the application.
• A vague business description (“consulting”) for a business that is actually e-commerce with inventory, or vice versa.
• Volume expectations that do not match the sector, in either direction.
• A residential address in a jurisdiction the institution avoids, with no context provided.
• Applying at the wrong tier: attempting a traditional bank remotely, or expecting a fintech to hold six-figure balances from day one.
• Filings already overdue at the moment of application: a Form 5472 problem is visible.
• A freshly issued EIN with no runway between issuance and application.
• Inconsistent KYC details: the legal name, its spelling, or the address does not match exactly across the passport, the Articles and the form itself.

The providers, one by one

“A US bank account” is not one product, and the providers are not interchangeable. Here is what each one actually is, who it fits, and where it breaks down in practice.

Mercury is a fintech, not a bank: it operates on top of partner banks (Choice Financial, Column N.A. and Patriot Bank among them), where your funds sit under FDIC coverage through those institutions. It fits tech, SaaS and e-commerce founders who value a polished product and clean integrations with Stripe and accounting software. It opens fully remotely with the EIN and the beneficial owner’s passport, no SSN or ITIN required, but it is unforgiving on business coherence and it restricts or closes accounts tied to certain higher-risk countries of residence. If your country sits on that list, applying to Mercury first can burn an attempt you did not need to spend.

Relay is also a fintech running on a partner bank, generally more flexible on residence than Mercury and built around splitting money into sub-accounts and giving team or bookkeeper access. Same document requirements (EIN plus passport, no SSN/ITIN) and it is frequently the right first move, not just the fallback, for founders whose country Mercury restricts.

Wise Business is not a bank either; it is a multi-currency payments account. It shines for invoicing and paying in multiple currencies at a competitive exchange rate, and it opens remotely with the same light documentation. Where it falls short is genuine US ACH routing in every configuration: some Stripe setups and some ACH collection flows do not sit cleanly on Wise alone, which is why it works best as a complement to Mercury or Relay rather than a sole account.

Traditional banks (Chase, Bank of America, Citi) are actual banks with branch networks and a durable banking relationship once you are in. They fit founders who live in, travel to, or plan to spend real time in the United States and want a long-term relationship that carries weight with lenders and enterprise counterparties. The cost is friction going in: an in-branch appointment is typical, and an SSN or ITIN is often requested from a non-resident. For a purely remote founder, this is rarely the realistic starting point: it becomes realistic once there is a reason to be there in person.

After a decline: how the retry works

A decline is not a verdict on you; more often it is a mismatch between your file and that specific institution’s risk appetite, or a fixable gap in the documentation. The response is the same discipline as the first application, not a scramble:

1. Identify the actual reason, where the institution states one. Vague inconsistency, sector risk, country restriction and missing documentation each point to a different fix.
2. Correct the specific gap (rewrite the business description, add the missing document, align the address) rather than resubmitting the same file hoping for a different reviewer.
3. Choose the right institution for the retry. Re-applying to the bank that just declined you, without changing anything, wastes the attempt. If the block was country-related, that usually means moving from Mercury to Relay, not trying Mercury again.
4. Wait a reasonable interval before reapplying rather than firing off three applications in a week: a pattern of rapid, repeated attempts is itself a signal that raises scrutiny.

Founders who treat the first decline as the final word give up on legitimate banking they could have secured with a corrected file. Founders who treat it as a blank check to spam every institution in sequence usually make the second and third attempts worse than the first.

KYC and AML signals that trigger review

Every institution runs some form of Know Your Customer and anti-money-laundering screening, and certain patterns reliably escalate a file from automated approval to manual review:

• A registered agent’s address used as the business address, with no separate operating address provided. It reads as a shell unless the description explains it.
• A country of residence flagged as higher risk by the institution’s own compliance policy: this is about their risk appetite, not a judgment on you, but it changes which provider is the right first attempt.
• A mismatch between the declared activity and the transaction pattern once the account is live: a “consulting” account suddenly moving high-volume e-commerce payments, for instance. This is the single most common trigger for a post-opening freeze, not just a pre-opening decline.
• Rapid, large initial deposits with no documentation trail explaining their origin.
• Ownership structures with unexplained layers: a holding entity behind the LLC that is not disclosed clearly on the application.

None of these are fatal on their own. They are the specific things a coherent file neutralises before they ever become a question.

The file we build

The deliverable is not a referral link. It is a dossier assembled once and reusable at every institution you approach:

• Formation documents: Articles of Organization from the state that fits your case, not the state with the best marketing.
• A real operating agreement, drafted for your ownership and activity, that reads like the business you described.
• EIN and the CP 575 letter from the IRS.
• A business summary written the way a compliance reviewer reads: what you sell, to whom, where they are, expected volumes, expected counterparties.
• Supporting evidence: client contracts, invoices, a working website, whatever substantiates the story.
• A bank shortlist for your specific profile, in the order that makes sense to approach them.

Then we walk the application with you, and if it comes back negative we regroup, correct and go again at the right institution. That loop is the product.

When traditional banking makes sense

Most non-resident founders start and stay with a fintech, and that is the right call while the priority is becoming operational fast. Traditional banking becomes worth the friction in a narrower set of cases: you travel to the US regularly or plan to relocate, you need business credit that a fintech account will not build toward, an enterprise client or a lender specifically expects a bank-of-record relationship, or your balances have grown to a point where FDIC coverage structure and in-person support start to matter.

What it actually requires from a non-resident, beyond the standard documents: usually an in-branch appointment (some banks will open with a US-based representative acting on your behalf, but that is the exception, not the rule), often an ITIN or SSN, and generally a more developed track record: an existing fintech account history, US invoicing history, or a US address of some kind strengthens the application considerably. Walking in cold, on day one of the LLC’s existence, with none of that behind you is the version of this that goes badly. The sequence that works for most founders is fintech first, traditional bank second, once there is a reason and a track record to bring to the branch.

If your account gets closed

It happens more often than the platforms admit, and it is rarely personal: it is usually the transaction pattern drifting from the stated business description, an automated risk model flagging volume or geography, or a periodic compliance review surfacing something that was never corrected. Fintech risk models in particular are automated and unforgiving: they close first and explain later, if at all.

The response mirrors the first application rather than requiring anything new: identify what likely triggered it (check for overdue state filings, a stale registered agent, or a business description that no longer matches what the account actually does), correct the underlying issue, and reapply at the institution that fits the corrected picture, sometimes the same one, more often a different tier. What makes this fast instead of catastrophic is having the documentation already assembled rather than reconstructing it under pressure while a client payment is stuck mid-transfer.

Stripe is a separate approval

Stripe is not a bank account and does not substitute for one: it is a payment processor. It collects from your customers and settles the proceeds into a bank account you already hold; without that account behind it, Stripe has nothing to settle into. It also runs its own, entirely independent risk review, separate from whichever bank or fintech holds your funds. A US LLC with a coherent business description and matching documentation generally clears Stripe’s review more easily than an inconsistent one, for exactly the same reasons a bank application does. Treat payment processing and banking as two separate approvals that both need preparing, not one bundled step.

When the answer is no

Sometimes the honest outcome is that US banking is not the right next step: because the residency question underneath is unresolved, because the activity sits outside every institution’s appetite, or because the structure needs fixing first. We say so. A declined application leaves a footprint, and burning three banks in a month is worse than waiting two weeks and applying once, properly.

If your structure came from a platform and you are not sure what state it is in, switching your LLC starts with exactly that audit.

In short

US banking for a non-resident LLC is decided by documentation, coherence and choosing the right institution, in that order. Fintechs open fast and remotely but close fast too; traditional banks are slow, sticky and worth the effort once there is a reason. The file is what you control, and it is what we build. What nobody controls is the decision itself. Anyone who tells you they do is not being straight with you.