FedNow vs RTP vs ACH: How America's Payment Rails Actually Work

Ask a payments engineer how long it takes to move money between two US bank accounts, and the honest answer is: it depends which rail you're on. It could be instant, or it could take three business days. That gap exists because the United States runs three separate systems for moving money electronically, built decades apart, operated by different institutions, with different rules about who can use them and what happens when something goes wrong.

FedNow, RTP, and ACH aren't competing versions of the same idea — they're different tools built for different jobs, and most banks now support more than one. Understanding how they actually work, not just their marketing taglines, matters for anyone building payroll, billing, marketplace payouts, or any product that touches US bank transfers.

What each rail actually is

ACH (Automated Clearing House) is the oldest of the three, dating back to the 1970s. It's a batch-processing network operated jointly by two organizations: The Clearing House (which runs the EPN operator) and the Federal Reserve (which runs FedACH). Transactions aren't sent one at a time — they're collected into batches and settled at scheduled windows throughout the day. ACH is the rail behind direct deposit paychecks, mortgage payments, utility bills, and most recurring subscription charges.

RTP (Real-Time Payments) launched in 2017, built and owned by The Clearing House — a private company owned by the largest US banks (JPMorgan Chase, Bank of America, Citi, Wells Fargo, and others). RTP was the first genuinely real-time, 24/7/365 payment rail in the US. Transactions settle individually, in seconds, with immediate finality — meaning once a payment clears, it can't be reversed by the sender.

FedNow is the newcomer, launched by the Federal Reserve in July 2023. It does functionally the same thing as RTP — instant, irrevocable, 24/7 payments between accounts — but it's operated by the central bank rather than a bank-owned consortium. FedNow exists partly because smaller banks and credit unions were wary of depending on a payment rail owned and governed by their largest competitors.

The core mechanical difference

The real dividing line isn't FedNow versus RTP versus ACH — it's batch versus real-time. ACH batches transactions and settles them at fixed windows. FedNow and RTP process and settle each transaction individually, the moment it's sent, with no waiting for a batch cutoff.

Why FedNow matters right now

FedNow's growth over its first two years is the clearest signal that instant payments are moving from novelty to infrastructure. In its fiscal year 2025, the value of payments processed on FedNow jumped more than 2,000%, reaching $853 billion, with roughly 1,800 financial institutions connected to the network. That's a sharp inflection from a standing start in mid-2023.

The scale of that jump matters for two reasons. First, it means FedNow crossed a credibility threshold — banks and credit unions that were waiting to see if the network would get real adoption before investing engineering time now have a much harder case to make for staying on the sidelines. Second, it signals that the use cases driving volume have shifted from experimental pilots to production traffic: payroll advances, insurance claim disbursements, real estate closings, and B2B settlement are increasingly routed through instant rails instead of waiting on ACH batch windows.

This is also a competitive story. RTP had a five-year head start and is directly owned by the largest US banks, giving it a natural distribution advantage with money-center institutions. FedNow's growth suggests the Federal Reserve's pitch to community banks and credit unions — a neutral, publicly operated alternative to a rail controlled by their biggest competitors — is working. Banks now increasingly connect to both networks rather than picking one, which is reshaping how payment processors and fintechs route transactions behind the scenes.

How money actually moves on each rail

It helps to walk through the mechanics rather than just compare speeds.

ACH step by step

1. The originating bank (or a payment processor acting as an Originating Depository Financial Institution) collects transactions throughout the day.
2. Batches are submitted to an ACH operator — either the Federal Reserve's FedACH or The Clearing House's EPN.
3. The operator sorts transactions by receiving bank and forwards them in batches at scheduled processing windows.
4. The receiving bank posts the funds, subject to its own funds-availability policies.
5. Standard ACH settles in one to three business days; Same Day ACH (available since 2016, with expanded windows since) can settle the same day if submitted before a cutoff, though same-day is not instant and still runs on scheduled windows rather than continuously.

RTP and FedNow step by step

1. The sending bank submits a single payment message the moment the sender authorizes it.
2. The network (RTP or FedNow) validates and routes the message to the receiving bank in seconds.
3. The receiving bank must accept or reject the payment immediately — there's no float period.
4. Once accepted, settlement is final. The receiving bank must make funds available to the account holder immediately.
5. This works continuously, including nights, weekends, and holidays — a structural change from ACH, which follows business-day banking hours.

The irrevocability point is worth dwelling on. With ACH, a transaction can be reversed — for fraud, insufficient funds, or an error — and consumer debits can be disputed for up to 60 days under Regulation E. With RTP and FedNow, once the receiving bank accepts the payment, it's done. This is why instant rails require much stronger fraud screening before the payment is sent, since there's no clawback safety net afterward.

Practical implications for businesses

The choice of rail isn't really a choice most businesses make directly — banks and payment processors decide which rails to support, and increasingly they support several, routing intelligently based on the transaction type. But knowing the differences still matters for anyone designing a payment flow.

• Payroll and gig payouts: Same-day ACH remains the default for standard payroll because it's cheap and predictable. Instant rails are increasingly used for on-demand pay features (workers cashing out earned wages before the normal pay date) and gig-economy payouts where speed is a competitive feature.
• B2B invoicing: ACH still dominates because most B2B relationships don't need instant settlement and ACH's lower cost per transaction matters at volume.
• Marketplace and platform payouts: Instant rails are increasingly attractive here — sellers and drivers who get paid immediately are measurably more satisfied, and platforms can charge a premium for "instant payout" as an upsell over free standard ACH.
• Bill pay and account-to-account transfers: FedNow and RTP support "Request for Payment" (RfP) messages, letting billers send a payment request that the payer can approve instantly — a mechanic that doesn't exist on ACH.
• Real estate and high-value transactions: The higher per-transaction limits on RTP (up to $1 million) make it more suitable than FedNow's lower default cap for large one-off transfers, unless the bank has separately negotiated a higher FedNow limit.

Cost considerations

ACH transactions are typically the cheapest per-transaction option, often a fraction of a cent to a few cents at the network level (though banks and processors mark this up for end customers). Instant payments on RTP and FedNow generally carry higher per-transaction fees, reflecting the real-time infrastructure and irrevocable settlement risk the receiving bank takes on. For high-volume, non-urgent transfers, ACH usually remains the economically rational default; instant rails earn their premium when speed itself is the product feature.

How businesses actually connect to these rails

Almost no business integrates with ACH, RTP, or FedNow directly the way a bank does. Direct membership in these networks requires being a financial institution or working through one, so most companies reach these rails through an intermediary layer.

• Banks and credit unions connect directly to the networks they support — often through a core banking provider or a network-certified gateway rather than building the connection themselves.
• Payment processors and fintechs (payroll platforms, payout providers, banking-as-a-service companies) maintain relationships with one or more sponsor banks, and route transactions across ACH, RTP, and FedNow behind a single API, often without the end business ever knowing which rail was used for a given transfer.
• Software platforms (marketplaces, gig apps, insurance carriers) typically sit a layer above that, using a processor's API and choosing between "standard" and "instant" payout options that the processor maps onto the underlying rail.

This layering matters because it explains why the rail-selection decision described earlier — same-day ACH for payroll, instant rails for on-demand payouts — is usually made by a processor's routing logic, not manually chosen per transaction. A well-built payout system will attempt an instant rail first when the receiving bank supports it, and automatically fall back to ACH when it doesn't, so the end user experience stays consistent even though the underlying path differs. Getting this fallback logic right is one of the more common integration challenges: teams need to handle partial network coverage gracefully rather than assuming every account can receive an instant payment.

For engineering teams evaluating processors, it's worth asking directly which networks a given provider supports, how they determine eligibility for instant payment on a receiving account, and what happens — both technically and from a customer-communication standpoint — when an instant payment attempt fails and falls back to ACH. Those details vary a lot between providers even though the marketing language ("instant payouts") tends to sound identical.

Limitations and open questions

Neither instant rail has universal reach yet. Not every US bank or credit union connects to FedNow or RTP — a business can't assume a given customer's bank supports instant receipt, which means most instant-payment products still need an ACH fallback for accounts that aren't reachable. That patchwork coverage is the single biggest practical constraint on adoption today, more than technology.

There's also a structural tension in having two competing instant-payment networks rather than one. Financial institutions that want to offer instant payments to all their customers may need to connect to both FedNow and RTP, since a given payment can only route to a bank that supports the same network the sender is using. This is different from ACH, where a single network reaches essentially every US bank account. Over time, this could push toward more banks joining both networks, but it adds real integration cost in the meantime.

Fraud is the other unresolved issue. Because RTP and FedNow payments are irrevocable, the entire cost of fraud prevention shifts to the moment before the payment is sent — there's no reversal mechanism to lean on afterward. Both networks have built fraud-monitoring tools (FedNow's is called FedNow FraudClassifier, and RTP has its own risk services), but the industry is still working out best practices, and instant payment fraud losses are watched closely by regulators.

Finally, consumer awareness lags bank capability. A bank can be technically connected to FedNow or RTP without offering instant payments as a visible feature to its retail customers — many banks have connected primarily for receiving payments, not yet for enabling customers to send them. The gap between "connected to the network" and "actively offering the product" is still closing.

What to watch next

The trajectory over the next few years will likely be shaped by a few forces:

1. Bank and credit union onboarding pace — how quickly the remaining large regional and community banks join FedNow and RTP, since coverage gaps are the main barrier to instant payments becoming a true default.
2. Request for Payment adoption — whether billers (utilities, insurers, subscription services) start using RfP messaging at scale, which would let consumers approve bills instantly instead of scheduling ACH debits.
3. Fraud tooling maturity — how effectively FraudClassifier-type tools and bank-side risk scoring reduce the irrevocability risk that currently makes some institutions cautious about enabling instant payments broadly.
4. Interoperability pressure — whether market or regulatory pressure pushes toward easier routing between FedNow and RTP, so senders and receivers don't need to be on the same specific network.
5. Same Day ACH evolution — the Fed and NACHA continue to expand Same Day ACH windows and limits, which narrows (but doesn't eliminate) the speed gap with instant rails for businesses that don't need true real-time settlement.

FAQ

Is FedNow the same as RTP?

No. They perform the same basic function — instant, irrevocable payments between bank accounts, available 24/7 — but they're separate networks. FedNow is operated by the Federal Reserve; RTP is operated by The Clearing House, a company owned by major US banks. A bank has to connect to each network separately.

Which is faster, FedNow, RTP, or ACH?

FedNow and RTP are both effectively instant, settling in seconds. Standard ACH takes one to three business days, while Same Day ACH can settle the same business day if submitted before a cutoff window — but neither ACH option is truly real-time or available outside business hours.

Can I choose which rail my payment uses?

Usually not directly. The rail is determined by your bank or payment processor based on what both the sending and receiving institutions support, and what the payment type requires. Some payment apps and platforms do let you pick "instant" (via RTP/FedNow, often for a fee) versus "standard" (via ACH, usually free or cheaper).

Is FedNow available at my bank?

Not necessarily. FedNow adoption is growing quickly, but not every US bank or credit union has connected yet. You'd need to check with your specific bank, and even if they've connected, they may only support receiving instant payments rather than sending them.

Why can't ACH payments be sent instantly?

ACH was designed as a batch-processing system from the start, collecting and settling transactions in scheduled windows rather than one at a time. Same Day ACH improved on this by adding more frequent windows within a business day, but it still isn't continuous, real-time processing the way FedNow and RTP are.

What happens if an instant payment is sent to the wrong account?

Because FedNow and RTP payments are irrevocable once accepted by the receiving bank, there's no automatic reversal mechanism. Recovery typically requires the sender to contact their bank, which then has to request the funds back from the receiving bank and the receiving customer — a process that isn't guaranteed to succeed, unlike ACH, where certain reversals are built into the network rules.

Do businesses pay more for instant payments?

Generally yes. RTP and FedNow transactions typically carry higher per-transaction fees than ACH, reflecting the real-time settlement and irrevocability risk involved. For high-volume, non-urgent transfers, ACH usually remains cheaper; businesses tend to reserve instant rails for payments where speed is a specific requirement or a paid feature.

Teams building payroll, payout, or billing systems that need to route intelligently across these rails can get hands-on help from Woyce Technologies.