- A check signing policy is a written document, separate from who your bank lists as an authorized signer.
- Banks generally do not enforce your dual signature rule. A check can clear with one signature even when your policy requires two.
- That makes the policy yours to monitor, which is the part most businesses miss.
- Set a threshold above which a second approval is required, and write it down.
- Nothing gets signed without supporting documents attached and every field already completed.
- Never pre-sign checks. It is the one rule that disables every other control at once.
- The signer checks the number sequence, which catches missing stock faster than reconciliation does.
- Small teams can still do this. Full separation is not always possible, but documented approval always is.
Most businesses have signers. Far fewer have a check signing policy, and the two are not the same thing.
Your bank knows who is allowed to sign. It does not know what your business requires before a check reaches that person, and it will not enforce anything you decide internally.
Here is what belongs in the written policy, the enforcement reality almost nobody leads with, and how to make it workable when you have four people rather than forty.
What is a check signing policy?
It is a written statement of how your business authorizes, prepares, signs and records outgoing checks.
It is not the same as signing authority. Authority is who your bank will accept a signature from, recorded on your signature card. Policy is everything that happens before a check reaches them.
| Signing authority | Check signing policy | |
|---|---|---|
| Set by | Your bank records | Your business |
| Answers | Who may sign | What must happen before they do |
| Enforced by | The bank | You. Nobody else |
| Exists as | A signature card | A written document |
For the authority side, including signature cards, entity types and what happens when someone leaves, see our guide to who signs a business check.
Does your bank enforce your check signing policy?
Generally no, and this is the most useful thing on this page.
Businesses routinely adopt a rule requiring two signatures above a certain amount, then assume the bank will stop anything that arrives with one.
It usually will not. A check bearing one valid signature can clear even when your own policy requires two, and banks typically do not accept responsibility for monitoring an internal rule they did not agree to.
So is a dual signature rule pointless?
No, but its value is different from what people assume.
A dual signature requirement does not work as a barrier at the bank. It works as an internal requirement that a second person looks at significant payments before they leave.
That second look is the control. The signature is just the evidence it happened.
Two practical consequences:
- You must monitor compliance yourself, usually during reconciliation
- Pre-signed blank checks defeat it entirely, because the second signature is applied before anyone has seen the payment
Our guide to writing a business check by hand covers the mechanics of signing, and whether signature stamps are legal covers delegation, which carries the same risk.
What should a check signing policy include?
Seven components, in the order they happen. A short written policy covering all seven beats a long one covering three.
Who may sign, and who may not?
State the roles rather than only the names, so the policy survives staff changes.
The most common exclusion is worth stating explicitly: whoever maintains the books generally should not also sign checks. In a very small business that may not be fully achievable, in which case say so in the policy and define the limits around it rather than leaving it unspoken.
Why must approval come before preparation?
This is the component most policies omit, and the one auditors ask about first.
A check should not be prepared until the payment itself has been approved. Many businesses approve at the signing stage, which is too late: the check already exists, and refusing to sign means voiding it.
A short approval record, whether a form, an initialed invoice or a documented email, establishes that someone authorized the spend before anything was printed.
What threshold should trigger a second approval?
Pick an amount above which a second person must approve or sign.
The right number depends on your business. Set it high enough that routine payments are not slowed, and low enough that a damaging payment cannot pass unnoticed. Review it as the business grows, because a threshold set at founding rarely still fits.
Write the number into the policy. A threshold everyone remembers differently is not a control.
What must be attached at signing?
The signer should receive the check together with whatever justifies it: the invoice, the approval record, the purchase order.
Signing without documentation is signing on trust, which is fine until it is not.
What should the signer actually check?
Specify it, because otherwise signing becomes automatic.
- The payee matches the invoice
- The amount matches the approval
- Every field is complete, with no blanks
- Funds are available
- The check number follows in sequence
The sequence check is the one nobody includes and the one a check printer would insist on. A signer who notices a number out of order has caught missing stock weeks before reconciliation would.
One thing the signer should not rely on
The memo line.
A memo describing a payment as a supplier invoice or a routine expense is written by whoever prepared the check. It is a label, not evidence. A memo that reads convincingly tells you only that someone typed it.
The same applies when reviewing statements later. Verification means matching the check against the underlying document, not against the description written on it.
What should the policy prohibit outright?
Short, absolute rules that need no judgment:
- Never sign a blank or incomplete check
- Never pre-sign checks for later use
- Never make a check payable to cash
- Never sign without supporting documentation
- Never leave signed checks unattended
On the third point, our guide to check safety covers why an untraceable payee is a problem.
Why is pre-signing the rule that matters most?
Because it is the gap most often exploited, and it usually starts as a convenience rather than a risk.
A signer travels, or is out for a week, and leaves a few signed blanks so payments are not held up. The intention is reasonable. The result is a negotiable instrument sitting in a drawer with no review attached to it.
Every other control in your policy assumes the signature comes last, after the payee, amount and documentation have been checked. Pre-signing inverts that order and disables the lot in one step.
If coverage during absence is the real problem, solve it as a coverage problem. Add a second authorized signer, agree that non-urgent payments wait, or arrange electronic payment for anything that genuinely cannot. Those are policy answers. A stack of signed blanks is not.
What happens after the check is signed?
The policy should not end at the signature.
A useful and frequently overlooked rule: the person who signs should be the one who sends. If a signed check returns to whoever prepared it, that person has an unsupervised window with a negotiable instrument.
Also cover how the payment is recorded, where the documentation is filed, and how voided checks are retained. Our guides to the check register and what is a voided check cover both.
How do you write this policy if you only have a few people?
The honest constraint, and the reason most small businesses skip the exercise.
Full separation of duties assumes enough people to divide the work. With four staff you cannot always have one person approve, another prepare, a third sign and a fourth reconcile.
What you can always do is document. Separation reduces opportunity. Documentation creates a trail. The second works at any size.
What adaptations actually work?
- Bring in an outsider for one step. An owner, a director or your accountant can review statements without daily involvement
- Lower the threshold instead of adding people. Fewer staff means more payments deserve a second look
- Allow a small exception and bound it. If the bookkeeper must sign in emergencies, set a strict limit and require after-the-fact review
- Separate the review rather than the signing. If one person must handle payments, someone else opens the statements
- Write the compromise down. A documented exception is a control. An undocumented one is a gap
That last point matters more than it looks. Auditors and accountants react far better to a stated, reasoned compromise than to silence.
How does the policy connect to your physical checks?
Most guidance on this subject stops at process. The document itself is a control surface too, and that part usually goes unwritten.
| Policy element | Physical control it depends on |
|---|---|
| Sequence checking at signing | Recording the check number range when stock arrives |
| Limiting who can prepare checks | Blank stock stored locked, access limited |
| Voided check handling | Voided checks filed rather than discarded |
| Detecting alteration after signing | Security features on the stock itself |
| Clean records for reconciliation | Format with a voucher or stub for detail |
The first row is the most actionable. Recording the number range on arrival takes seconds and makes sequence checking possible. Without it, "the numbers look right" means nothing.
For the stock side, see high-security business checks and check security features. For formats, see types of business checks, computer checks, manual business checks or payroll checks.
What is this policy actually protecting against?
Worth naming, because a policy without a stated purpose gets treated as bureaucracy.
- Payments to fictitious payees, caught by requiring supporting documentation
- Inflated amounts, caught by matching the check to the approval
- Stock taken from the back of a book, caught by sequence checking
- Unauthorized spending, caught by approval before preparation
- Simple errors, caught by a second pair of eyes
Note how many involve someone with legitimate access. Security paper defends against outsiders. This policy is the defense against everyone else, and there is no product that substitutes for it.
Our guide to check kiting fraud covers how internal access can be misused, including through company accounts.
How often should you review the policy?
Annually as a minimum, and immediately whenever any of the following happens:
- Someone with signing authority joins or leaves
- The business grows enough that the threshold no longer fits
- A new bank account is opened
- Payment volume changes materially
- An incident or near miss occurs
- An accountant or auditor raises it
Point one is the one that slips. Policies list people who left months ago, and bank records are updated long after the office has moved on.
Should any review be unannounced?
Worth considering, and it adds nothing to your workload.
A review that happens on a known schedule is predictable, and predictability is something anyone misusing the process can plan around. An occasional unscheduled look at the register, the statements or the check stock behaves differently, because it cannot be prepared for.
This is not about suspicion. It is about making the control real rather than ceremonial, and a policy that says reviews may happen at any time is easier to apply evenly than one that singles anyone out.
What should prompt a closer look?
Some patterns are worth noticing regardless of who is involved, and a policy is easier to enforce when it names them in advance rather than reacting to a person.
- Reluctance to hand over payment duties, including during holidays or illness
- Resistance to anyone else reviewing statements
- Records that are always about to be ready rather than available
- Gaps in the check sequence with no voided check filed
- Statements arriving already opened by whoever handles payments
None of these proves anything alone. They are reasons to apply the policy you already wrote, which is precisely why having written it matters.
What we see from businesses that write checks
We print and ship business and personal checks for U.S. customers every business day, and payment controls come up constantly at order time.
Businesses believe the bank enforces their two-signature rule. It is the most consequential misunderstanding here, and most owners are surprised to learn otherwise.
Nobody records the check number range on arrival. It takes seconds, it is the foundation of sequence checking, and it is the first thing needed when something goes missing.
Approval happens at signing rather than before. By then the check exists, and declining means voiding rather than simply not proceeding.
Blank stock sits unlocked. Almost universally, and often alongside whatever is used to sign.
The policy exists only in someone's head. It works while that person is there and disappears the week they are not.
Putting your check signing policy in writing
A check signing policy is your document, not your bank's. Banks record who may sign and generally will not enforce a dual signature rule or any other internal requirement, which makes monitoring your responsibility.
Cover the seven components: who may sign, approval before preparation, the threshold for a second approval, what must be attached, what the signer verifies, what is prohibited, and what happens afterward. Keep it short enough that people read it.
If your team is small, document the compromise rather than skipping the policy. A written, reasoned exception is a control. Silence is a gap.
This is general information rather than legal, audit or accounting advice, and requirements vary by business and industry. Your accountant is the right person to review anything specific to your situation.
Reviewing your controls? Browse business checks or high-security business checks, or contact us if you want help matching stock to how your business handles payments.