Opening a US bank account as a non-resident, honestly

Almost everything written about this is a ranked list of providers, which is the wrong shape for the problem. Whether you can open a US business account has very little to do with which brand you prefer and almost everything to do with four facts about you: your nationality, your country of residence, what you sell, and whether you hold a US tax identification number.

Two founders with identical Wyoming LLCs, applying on the same day to the same provider, routinely get opposite answers. Nothing about the company explains it.

So this is a decision tree instead. Work down it, and by the end you should know which door is actually open to you and what to do if it closes.

First, the distinction that saves you weeks

"US bank account" describes two genuinely different products, and conflating them is why people spend months failing at the wrong one.

For a non-resident founder with a US LLC and no US presence, the honest answer is that the fintech route is the route. Traditional banks are not impossible, but they generally expect you in a branch with your passport, and the ones that don't tend to want a relationship, a balance, or a US-resident officer.

If you need to deposit cash or receive paper cheques, you need a traditional bank and should plan a trip. If you need to receive wires and card payments and pay vendors — which is what most founders actually need — you do not.

The decision tree

1. Do you have an EIN?

If not, stop here. The EIN is the company's federal tax identification number and it is the one document no provider will proceed without. As a foreign owner with no SSN or ITIN, you obtain it by submitting Form SS-4 rather than through the online application, which is slower — budget accordingly, because everything downstream waits on it.

2. Do you hold an SSN or ITIN?

This is the fork that has moved most recently, and it is worth being precise about.

A personal US tax identification number is not universally required. Several providers verify a non-resident owner using a passport and proof of address, relying on the company's EIN as the tax identifier. Mercury is the most prominent example — it has been openable by non-residents without an SSN or ITIN.

But this is not a stable industry norm. Relay moved to requiring an SSN or ITIN, which removed it as an option for exactly the founders who previously used it. Assume any provider can make the same change, and treat "no ITIN needed" as a current fact about one company rather than a property of the category.

3. What passport do you hold, and where do you live?

This is the factor that decides more applications than any other, and the one least discussed, because it is uncomfortable.

Every US-facing provider enforces OFAC sanctions, which is not discretionary. Comprehensively sanctioned jurisdictions — North Korea, Iran, Syria, Cuba and others — are categorically out. Russia and Belarus are refused across the major fintechs regardless of where the company is registered.

Beyond the sanctions floor, each provider maintains its own risk-based restricted list, which is longer, unpublished in full, and revised regularly. Nationality and country of residence are assessed separately, so holding an acceptable passport while residing somewhere restricted can still produce a decline.

A Wyoming LLC does not launder this. The screening is on the beneficial owner, not the entity — which is the entire point of the customer due diligence rules that require providers to identify and verify the individuals behind a legal entity customer at account opening.

4. What do you actually sell?

Industry screening is the second-largest cause of refusal and the easiest to trip accidentally. Categories that reliably draw scrutiny or outright decline:

• Anything crypto-adjacent, including businesses that merely accept crypto payment.
• Gambling, betting and most gaming with a cash-out mechanic.
• Adult content and adult-adjacent platforms, including agencies representing creators.
• Money services, remittance, and anything that looks like holding customer funds.
• Pharmaceuticals, supplements and CBD.
• Multi-level marketing, lead generation and high-refund-rate e-commerce.

Describe your business accurately at application. Vague descriptions read as evasive and get declined; inaccurate ones get the account frozen later, after money is in it, which is materially worse.

What the application actually asks for

Assemble this before you start. Applications time out, and re-submitting after a gap reads worse than applying once, completely.

• Formation documents — the filed articles or certificate of formation from the state.
• The EIN confirmation letter — the CP-575, or a 147C replacement if you have lost it.
• The operating agreement, showing ownership and control. Single-member LLCs need one even though most states do not require it.
• Passport for every beneficial owner, and usually for anyone with control.
• Proof of personal address — a utility bill or bank statement, typically within 90 days, matching the address you give.
• A US business address for the company. A registered agent address is sometimes accepted and sometimes rejected; a virtual address from a mail-handling provider is more often accepted, but not universally.
• A clear description of the business, its customers, and expected monthly volumes.

Payments are a separate problem

An account that receives wires is not the same as the ability to charge customers, and founders regularly solve one and assume the other.

A US LLC with an EIN and a US business account is generally enough to onboard with the major processors. What comes after onboarding is where non-resident founders get hurt:

• Rolling reserves. A percentage of your volume held for a set period, imposed at the processor's discretion based on your risk profile. A new entity with a foreign owner in a high-refund category is a textbook candidate.
• Sudden review holds. A volume spike — a launch, a good month — can trigger a review that pauses payouts precisely when you need them.
• Termination without a stated reason. Processor agreements generally permit it, and appeals are slow.

Open the second payment rail while you are healthy and boring. Applying for one during a freeze, with a frozen account as your reference, is the worst possible time to be assessed.

Concretely: two banking relationships, two processors, and enough operating cash outside your primary rail to run for a month. This is not paranoia — it is the standard failure mode for exactly this profile of business, and the mitigation costs an afternoon.

When you are declined

You will usually be given no reason, and asking will not produce one. Do this instead, in order.

1. 1Do not immediately reapply to the same provider. A second decline within days confirms the first rather than overturning it. Wait, and change something material first.
2. 2Audit the coherent-story problem. Reconcile your addresses, the business description and the ownership documents. This fixes a genuine share of declines with no change to your circumstances.
3. 3Try a structurally different provider. If a US-centric fintech declined you, a provider built around international account holders is a different underwriting model, not just a different brand.
4. 4Consider whether the ITIN unlocks it. If the decline came from a provider that requires one, this is the whole answer, and it is a matter of months.
5. 5Reassess the entity. If your nationality is the blocker for US-facing providers generally, a US LLC may be the wrong container. That is a structuring question, not a banking one.

The fifth point is the one worth sitting with. Founders sometimes spend six months failing to bank a US LLC that was never going to work for their passport, when the underlying business would have banked cleanly elsewhere. The company should follow the constraint, not fight it.

The short version

The uncomfortable summary is that this process is not merit-based and does not respond to persistence. It is a set of screens, most of them about facts you cannot change quickly. Knowing which screen you are failing is worth more than another application — and the honest version of that information is what almost nobody publishes, because it does not convert as well as a list of ten best accounts.