A timely filing limit is the deadline a payer sets for receiving a claim after the date of service. Miss it, and the claim is denied for timely filing, usually with code CO-29, even if the care was covered, the coding was correct, and the documentation was complete. That is what makes this denial different from most others. A coding denial can be corrected. A timely filing denial usually cannot, because the clock has already run out.
Every payer runs its own clock. Original Medicare gives 12 months. Commercial contracts commonly give 90 to 180 days. Medicaid varies by state, inside a 12 month federal ceiling. The contract or payer manual for that specific claim controls, and any national table, including the one in this post, is a starting point for checking, not a substitute for the contract.
For where a claim sits in the wider cycle, see how medical billing works. For what happens when a claim is refused after review, see claim denials in medical billing: common causes and fixes.
What a timely filing limit is
A timely filing limit, often shortened to TFL, is the maximum time between the date of service and the date the payer receives a clean claim. Three details in that definition do most of the work.
First, the date that counts is the payer receipt date, not the date the practice sent the claim. A claim mailed on the deadline that arrives three days later is late. Electronic submission through a clearinghouse removes most of this ambiguity, because the clearinghouse acceptance report and the payer acknowledgment show when the claim was received. Keep those reports. They are the proof of timely filing if a payer later says a claim arrived late.
Second, the claim has to be clean, meaning accepted into the payer system. A claim that is rejected for missing or invalid data never entered adjudication, so the payer treats it as not received at all. A claim that is denied, by contrast, was accepted and then adjudicated, and the denial itself starts its own separate deadlines for correction or appeal. Rejected, denied, and corrected are three different states, and confusing them is one of the common ways a claim drifts past its deadline.
Third, the limit is a contract term, not a courtesy window. Payers are not obliged to make exceptions for delays within the practice's control, such as a claim sitting unbilled or a work queue backlog. Exceptions exist, but they are narrow, and they are covered below.
When the clock starts
The clock does not always start on the same date. The start date depends on the claim type, and getting it wrong by even a few weeks can cost a claim.
| Claim type | When the clock usually starts |
|---|---|
| Professional claim | Date of service |
| Inpatient facility claim | Date of discharge, the "through" date on a span claim |
| Secondary or coordination of benefits claim | Often the date of the primary payer remittance, for many commercial payers, but not for Original Medicare as secondary |
| Corrected claim | Often measured from the original remittance date, under that payer rule, not from a fresh date of service |
| Span claim | Varies for span claims; check the CMS span-date rule for the claim type |
Two of these deserve emphasis. For secondary claims, many commercial payers restart the window from the primary payer remittance, with a window such as 90 days from that date. Original Medicare does not restart: as a secondary payer it still requires the claim within 12 months of the date of service, even if the primary payer took months to pay. A billing team that assumes every secondary clock restarts will lose Medicare secondary claims.
For corrected claims, a correction does not reset the original filing window. Many payers measure the corrected claim window from the original remittance date, or allow a set number of days from payment or denial, whichever is later. A claim filed on time and then denied for a correctable reason still has to be corrected inside that second window. The original timely filing does not buy unlimited time to fix it.
What the limits look like by payer type
The table below gives the shape of the limits, with the sources a practice should check for each row. Treat every commercial and Medicaid figure as a range to verify, not a fixed rule.
| Payer type | Typical initial limit | Controlling source |
|---|---|---|
| Original Medicare, Part A and Part B | 12 months from the date of service, or discharge for inpatient care | 42 CFR 424.44 and the CMS Claims Processing Manual, Chapter 1, Section 70 |
| Medicare Advantage | Set by the plan contract, commonly 90 to 180 days for contracted providers | Provider agreement and plan manual for that plan |
| Medicaid fee for service | Federal ceiling of 12 months, with individual states setting shorter limits | State Medicaid manual |
| Medicaid managed care | Set by the state contract and plan manual, varies by plan | Managed care plan manual |
| Commercial insurance | Commonly 90 to 180 days for in-network claims, with out-of-network windows sometimes different | Provider contract and payer manual |
| TRICARE and CHAMPVA | Generally 12 months from the date of service, or discharge for inpatient care | TRICARE: 32 CFR 199.7(d); CHAMPVA: 38 CFR 17.276; check the program regulation for the claim |
Three cautions belong with this table. First, network status matters. Out-of-network claims frequently carry different, sometimes shorter, windows than in-network claims with the same payer. Second, plan type matters. A payer name alone does not settle the deadline, because the same carrier runs commercial, Medicare Advantage, and Medicaid plans under different contracts. Third, contracts change. Payers revise filing limits with notice, and a matrix built last year can quietly go stale. When a payer changes a limit mid-contract, update the payer matrix and re-run the aging report in the same week.
What happens when a claim misses the window
A late claim is denied with Claim Adjustment Reason Code 29, which the X12 standard defines as "the time limit for filing has expired." On a remittance it shows up as CO-29. The CO group code means contractual obligation, which is why the balance stays with the practice rather than moving to the patient on an in-network claim. Most participation agreements prohibit billing the patient for a balance denied because the provider filed late, and Medicare rules prohibit it as well. In practice the balance becomes a write-off, unless a documented exception applies.
Appeal rights are limited, and they differ by payer. Under Original Medicare, a claim denied for exceeding the time limit is not considered an initial determination, because it is stopped before coverage is adjudicated. That means it is not eligible for redetermination, the first level of appeal. Providers instead pursue a reopening, and only when a valid exception applies. Commercial and Medicaid appeal rules are set by contract and plan manual, and they often give a short window from the remittance date, so a timely filing denial is a same-week task, not a monthly cleanup item.
It is worth separating this from other denials. For the code groups and what each one means, see common claim denial codes and what they mean. A timely filing denial differs from most other denials in that it is about the calendar rather than the care, the code, or the documentation.
The narrow exceptions
Exceptions to timely filing limits exist, but they are narrow, and they require documentation. The Medicare exceptions are the clearest example, and they illustrate the pattern other payers follow.
Under Medicare rules, the recognized exception categories include administrative error by a Medicare contractor or government employee, retroactive Medicare entitlement, a Medicare Secondary Payer situation, and a state Medicaid agency recouping funds from a provider six months or more after the fact. Even then, the extension is limited: the claim generally must be filed by the last day of the sixth calendar month following the month in which the error was corrected or the entitlement was established, and under 42 CFR 424.44 and the CMS Claims Processing Manual, Chapter 1, Section 70, the window cannot be extended beyond December 31 of the third calendar year after the year services were furnished, for the qualifying error categories.
An exception request stands or falls on its documentation. It needs a written explanation, the original claim details, and documentation showing the qualifying circumstance and its dates. A claim that was simply filed late because the practice was busy, short staffed, or waiting on records does not fit any of these categories. Practices that want an exception approved should treat the request as a small case file, not a phone call.
Rejected is not filed, and other distinctions that cost claims
Several distinctions in the claim lifecycle look minor and decide whether a claim survives. Each one is worth stating plainly.
A rejected claim was never filed. A rejection happens before adjudication, usually for missing or invalid data, and the claim does not appear on a remittance advice. The fix is to correct the data and resubmit as a new claim, inside the original window. A 999 acknowledgment only shows a file passed format checks at the clearinghouse stage; it is not proof the payer accepted the claim. The stronger proof is the payer 277CA acceptance or a payer claim number.
A denied claim was filed. A denial happens after adjudication, appears on a remittance advice, and carries its own correction or appeal deadlines. The fix depends on the denial reason, and the corrected claim often uses frequency code 7 to replace the original claim, while a void uses frequency code 8. A rejected claim that goes back as a new claim uses frequency code 1. Using the wrong frequency type can create a duplicate or a second denial, so the claim status on the remittance should decide the resubmission type, not habit.
A corrected claim is not a new filing window. As noted above, many payers measure corrected claim deadlines from the original remittance date. A claim filed on day 80 of a 90 day window, denied on day 100, and corrected on day 170 may already be outside the corrected claim window even though the original filing was timely. The corrected claim clock deserves its own column in the payer matrix.
A clearinghouse acceptance is not a payer acceptance. The clearinghouse routes and scrubs the claim, and its report shows the claim left the practice system in good order. The payer acknowledgment shows the payer received it. For proof of timely filing, keep both, and treat the payer acknowledgment as the one that settles a dispute. For what a clearinghouse does between a practice and its payers, see what a medical billing clearinghouse does.
How practices stay inside the window
No single tool prevents timely filing denials. What works is a small system, run on a short cycle, with the payer matrix at its center.
Build a payer matrix. One row per payer and plan type, with columns for the initial filing limit, the clock start date, the corrected claim window, the secondary claim window, and the appeal deadline. Fill it from the provider contract and the current payer manual, not from a blog post or a colleague memory. When a contract renews or a payer notice arrives, update the matrix first, before the next billing run.
Work the aging report by deadline, not by balance. Sort unbilled claims and rejected claims by date of service, oldest first, and work the shortest-window payers first. A claim for a 90 day payer that is 60 days old is more urgent than a larger Medicare claim that is 60 days old with 10 months left. Aging buckets that only show 30, 60, and 90 days past due hide this, because a claim can be inside the aging bucket and already past its payer deadline. Add a "days to deadline" column calculated from the payer matrix.
Work rejections within days, not weeks. A rejected claim is an unfiled claim with the clock running. A daily or twice-weekly rejection queue, cleared the same day it appears, removes a common source of preventable timely filing losses. The same applies to missing information denials: every day spent waiting for a record is a day off the corrected claim window.
Track proof of filing as a routine step. File the clearinghouse acceptance and the payer acknowledgment with the claim record, so a payer dispute about receipt can be answered with a document rather than a reconstruction. When a payer says a claim arrived late, the acknowledgment date is the answer.
Watch the handoffs that create late claims. Credentialing delays are a known trap: if credentialing takes longer than a payer filing window, claims for services in that period can expire before the provider is even enrolled. Eligibility checks that are skipped or run late create the same problem downstream, because an eligibility denial discovered after filing starts a correction cycle against a shrinking window. For the eligibility side, see how medical billing works for the claim lifecycle in order.
Common mistakes, and the fix for each
| Mistake | What goes wrong | The fix |
|---|---|---|
| Treating the submission date as the filing date | A claim sent on the deadline but received after it is late | Count from the payer receipt date, and use electronic submission with acknowledgments |
| Treating a rejected claim as filed | The clock keeps running while the claim sits in a queue | Work rejections within days, and resubmit as a new claim inside the original window |
| Assuming the secondary clock always restarts | Original Medicare as secondary still uses 12 months from the date of service | Check the secondary payer rule, and file Medicare secondary claims early even when the primary is slow |
| Assuming one deadline per payer name | Plan type, network status, and state change the window | Build the matrix by payer and plan type, not by payer name alone |
| Using a stale payer matrix | A revised limit quietly expires claims that used to be safe | Update the matrix on every contract renewal and payer notice, and re-run aging the same week |
| Letting a denied claim sit until month end | The corrected claim window can close before the monthly cleanup | Treat timely filing and corrected claim deadlines as same-week tasks |
| Billing the patient for a late claim balance | In-network contracts generally prohibit it, and it creates a compliance problem | Write the balance off under the contract rule, and fix the workflow that caused the late filing |
When to get help with timely filing
Timely filing problems usually show up as a pattern, not a single claim. If CO-29 denials cluster around one payer, one plan type, or one stage of the workflow, such as rejections or secondary claims, the cluster points to the fix: a payer rule missing from the matrix, a rejection queue that runs too slowly, or a handoff that adds weeks before the claim is first submitted.
Practices in that position often start with two reviews. First, pull every CO-29 denial from the last six months and sort by payer, plan type, and the stage where the delay happened. Second, compare the payer matrix against the current contracts for the payers in that list. The gap between the matrix and the contract, or between the contract and the actual work queue timing, is usually where the claims were lost.
The underlying rule stays simple. A timely filing limit is a calendar rule with almost no forgiveness, so the system has to catch a claim before the deadline, not after it. Know the deadline for each payer and plan, know when each clock starts, keep proof that the payer received the claim, and work rejections and corrections on a short cycle. Claims that are filed inside the window can still be denied for other reasons, but at least they get adjudicated. Claims filed outside it usually do not get that chance.
For firms that handle revenue cycle work, including denial management and accounts receivable follow-up, see our denial management services and revenue cycle outsourcing services. Our methodology explains how directory listings are vetted, including the services and scope firms claim.