Subcontractor Retainage Release Schedule Tracking Guide

A subcontractor retainage release schedule is the per-sub, per-project workpaper a general contractor's controller uses to fire retainage releases on a state-law-conditional clock. Each row encodes the state retainage cap that governs the project, the trigger rule (substantial completion versus final acceptance), the prompt-payment statutory clock that overlays the GC's pay-when-paid posture, and a pre-condition gating list that has to flip to satisfied before AP cuts the check: final unconditional lien waiver from the sub, lower-tier waivers from sub-subs and suppliers, surety consent on bonded work, final WH-347 on Davis-Bacon work, current COI, warranty acknowledgement, and 1099-NEC reconciliation closure on year-end-spanning releases.

The reason this workpaper is changing right now is California. California's SB 61 5 percent retention cap on private works adds Section 8811 to the Civil Code, prohibiting any retention payment withheld from a payment for a private work of improvement from exceeding 5 percent of the payment and capping total retention at 5 percent of the contract price, applicable to contracts entered into on or after January 1, 2026. For a GC operating in California, that cap lands as a column on the tracker — a 5 percent ceiling that overrides whatever percentage the prime or subcontract template has carried for the last decade. New York's 2023 5 Percent Retainage Law and Georgia's 2022 public-works rule have already done the same thing on their sides. The release-schedule workpaper is where those rule shifts become operational.

This article covers the post-substantial-completion release schedule — the workpaper that fires releases — not the per-pay-app retainage withholding line on the cover sheet, which is its own discipline and is covered separately in the article on G702 and G703 cover-sheet retainage withholding on each pay app.

The state-law rule landscape as workflow input

The federal floor is FAR 52.232-5: on federal-prime construction contracts the contracting officer may withhold up to 10 percent for unsatisfactory progress, with release as performance becomes satisfactory. Federal-project rows carry a 10 percent ceiling and a discretion-of-CO trigger.

California is two rules in one column. Public Contract Code §7107 governs public works and the prompt-release timing once the project is accepted. Civil Code §8811, added by SB 61, caps private-work retention at 5 percent of each progress payment and 5 percent of the contract price for contracts entered on or after January 1, 2026. The California rows split by project type and by contract date — public follows §7107, pre-2026 private follows the legacy cap, post-2026 private locks at 5 percent.

New York's 5 Percent Retainage Law (2023, amending General Business Law §756 and the Lien Law) caps retention at 5 percent on private contracts of $150,000 or more, requires release of subcontractor retention within 30 days after substantial completion of the sub's work, and within 7 days of the GC's receipt from the owner. The New York column carries a 5 percent cap and the 30-day post-substantial-completion sub clock.

Florida splits public and private. §218.735 caps public-project retention at 5 percent on contracts of $200,000 or more with release timeframes after substantial completion; Chapter 713 governs private-work liens and shapes the retainage-and-lien interaction. Georgia's OCGA §36-91-50 caps state and local public works at 5 percent (updated 2022).

Texas governs through Property Code Chapter 28 (Prompt Payment Act, public works) and Chapter 53 (private liens), with a 7-day GC-to-sub clock. Alabama §39-2-12 is structurally different: 5 percent through 50 percent completion, then no further withholding — a percent-of-completion shutoff column on the tracker, not just a cap.

The remaining 40-plus states follow variants of these patterns. The ASA Georgia 50-State Retainage Law summary is the canonical comparative reference, and our state-by-state retainage cap landscape and AP tracking guide carries the broader cap-and-rule grid. State statutes move — CA SB 61 (January 1, 2026), NY 2023, GA 2022 are the recent anchors — and the controller refreshes the rule reference at least annually.

Each named rule lands in the project tab as one cell across five columns: state, project type, retainage cap rule, release trigger rule, prompt-payment clock. The trigger rule, the clock, and the gating list that gates them are what the next sections walk.

Two-stage release: substantial completion and final acceptance

AIA A201–2017 §9.8 defines substantial completion as the stage at which the work is sufficiently complete in accordance with the contract documents that the owner can occupy or use the project for its intended purpose. §9.10 sets out final completion and final payment — the architect's final certificate, the contractor's affidavit that all payrolls and indebtedness have been paid, and the release of the remaining retainage. The contract recognises two distinct events; the release schedule has to do the same.

The first event is the substantial-completion release. Under most modern prime contracts and a majority of state-law postures, the owner reduces retainage at substantial completion and releases the bulk of what's been withheld, retaining only a punch-list-cost holdback against the work that remains. The standard sizing is 150 to 200 percent of the architect's or owner's estimate of the cost to complete — sized that way because punch-list costs frequently overrun the initial estimate and the holdback has to cover the overrun without forcing the owner back for a true-up.

The second event is the final-acceptance release. After the punch list is complete, the architect or owner signs off on final acceptance, and the remaining holdback releases. On state and contract postures that permit substantial-completion reduction, the second event is small — only the punch-list holdback. On postures that don't permit reduction, the second event is the entire retainage balance. The tracker's release trigger rule column on the project tab encodes which posture applies: substantial-completion-permits-reduction releases bulk-then-holdback; final-acceptance-only collapses both events into one release at final.

The downstream subcontract layer is not automatically aligned with the prime. When AIA A201 governs the prime contract and reduction is permitted at substantial completion, the GC's downstream subcontracts follow suit only if each subcontract incorporates A201 by reference or carries its own substantial-completion-reduction clause. Subcontracts that are silent on substantial-completion reduction may give the sub a contractual claim to no reduction until final acceptance — meaning the GC receives bulk retainage from the owner at substantial completion but cannot pass a corresponding release to a sub whose contract doesn't permit it. This is the most common reason a tracker's substantial-completion-release column needs to be a per-subcontract value rather than a per-project default. Read each subcontract's payment and final-payment clauses on award and set the column at that point, not at substantial completion when the timing is no longer flexible.

Each subcontractor row on the tracker therefore carries two target release dates and two target amounts: a substantial-completion release date with a substantial-completion release amount (and the corresponding punch-list-cost holdback held against final), and a final-acceptance release date with the holdback amount. The ledger has to record both events; the next section walks how.

Retainage receivable and retainage payable: the two-sided ledger

Every withholding event on a pay app accrues retainage on both sides of the GC's books. Retainage receivable accrues when the owner withholds retainage from the GC's pay app — typically posted to a separate accounts-receivable subaccount or, in some chart-of-accounts conventions, to a contra-revenue account that nets against billed revenue until released. Retainage payable accrues when the GC withholds retainage from the subcontractor's pay app and is posted to a separate liability account on the GC's balance sheet, distinct from regular trade payables. Treating both as discrete ledger lines (rather than netted into AR and AP) is what makes the release-schedule workpaper reconcilable against the GL.

The four canonical journal events drive the ledger:

• Pay app issued, owner withholds retainage from the GC. GC records the gross billing as revenue, debits regular AR for the cash portion of the pay app, and debits retainage receivable for the withheld portion.
• Pay app issued, GC withholds retainage from the sub. GC records the sub's billed amount as a payable, credits regular AP for the cash portion, and credits retainage payable for the withheld portion.
• Owner releases retainage to the GC. Cash debits, retainage receivable credits — the receivable is settled.
• GC releases retainage to the sub. Retainage payable debits, cash credits — the liability is settled.

The two accruals are linked but not equal. Retainage receivable is sized by the prime contract and the owner's withholding rule for the project (typically the state cap on owner-side retention); retainage payable is sized by the aggregate of subcontract terms, each subject to whatever cap the state imposes on subcontractor-side retention and whatever percent each subcontract negotiated within that cap. On a project where the owner is withholding 5 percent and the sum of the GC's subcontracts withhold a blended average of 7 percent, the payable balance will exceed the receivable balance — and that mismatch is normal, not an error. The number it should equal is the per-row payable column on the tracker, not the receivable column.

Reconcile monthly. The retainage payable subledger should tie out to the sum of the outstanding-retainage column across every sub row of the release-schedule workpaper for the month-end date. The retainage receivable subledger should tie out to the project tab's owner-retainage-outstanding cell summed across all active projects. Reconciliation breaks usually point to one of three things: a pay-app retainage line posted to the wrong account (regular AP instead of retainage payable, or vice versa), a release event recorded on the GL but not flipped on the tracker, or a substantial-completion release recorded as a single event when it should have been split between the bulk release and the punch-list-cost holdback. The schedule is the operational view of where every dollar sits and when each release fires; the GL is the audit trail. They have to agree on the closing date of every period.

---

Pay-when-paid posture, layered with state prompt-payment clocks

Pay-when-paid is the standard contractual posture in a subcontract: the GC releases retainage to the subcontractor only after receiving the corresponding retainage from the owner. Read narrowly, it's a timing mechanism — it sequences the GC's payment obligation to follow the owner's payment to the GC. Pay-if-paid is the more aggressive variant; it attempts to shift the entire risk of owner non-payment to the sub, making the owner's payment a condition precedent rather than a timing trigger. California, New York, North Carolina, and several other states treat pay-if-paid clauses as void against public policy, leaving pay-when-paid as the operative posture even where contracts attempt the stronger language. The tracker doesn't need to encode the doctrinal distinction, but the controller has to know which it is — when the owner is delinquent, the question of whether the sub still has to be paid is governed by which clause the contract actually uses and what the governing state will enforce.

State prompt-payment statutes overlay pay-when-paid with a statutory clock. Once the GC actually receives retainage from the owner, the GC has a defined window — usually 7 to 15 days — to pass that retainage down to the sub. The window varies by state and by project type. New York's 5 Percent Retainage Law requires release of subcontractor retention within 30 days of substantial completion of the sub's work and within seven days of the GC's receipt of retainage from the owner on private contracts of $150,000 or more. Texas's Prompt Payment Act sets the GC-to-sub clock at 7 days after receipt. Florida §218.735 sets a 10-day clock on public projects. California's §7107 sets specific clocks on public works and is supplemented on the private side by the SB 61 cap and the existing prompt-payment provisions of the Civil Code.

Where the two layers conflict, the statutory clock wins once it starts. Pay-when-paid lets the GC defer release indefinitely while waiting for the owner — but the moment the owner pays, the prompt-payment statute caps the deferral at the statutory window. The clock typically does not start until the conditions precedent are satisfied, which means the gating columns are time-critical, not just compliance-critical. A complete gating list on the day the owner pays starts the clock immediately and forces release within the statutory window. An incomplete gating list either delays the clock's start (in jurisdictions that read conditions precedent as gating the prompt-payment trigger) or arguably runs the clock anyway with the GC exposed to a statutory penalty for missing it. Treating the gating list as something AP can chase down after the owner pays is the wrong sequence; the gates have to be green when the owner pays, not after.

The tracker's release-target date column on each sub row is therefore computed from the date the GC receives retainage from the owner plus the state's prompt-payment clock — gated by completion of the per-sub pre-conditions. A sub whose final unconditional waiver and lower-tier waivers are not in hand on the day the owner pays the GC starts the clock late or, depending on jurisdictional reading, exposes the GC regardless. The next section walks the gating columns themselves.

The pre-condition gating list as tracker columns

Before AP releases a sub's retainage, every gate has to flip to satisfied. Each gate is a column on the sub row of the tracker; the release-target date computes only when every column is green. Treat the columns as hard gates rather than as compliance reminders the AP team will chase down once the release is already in motion.

Final unconditional lien waiver from the sub. The form must be the unconditional final type — not a conditional final, not an unconditional progress, but the version that takes effect on signature without contingency on payment receipt. Several states (California, Florida, Texas, Wyoming, Utah, Mississippi, Arizona, Nevada among them) prescribe statutory forms; a non-conforming waiver in a statutory-form state is unenforceable. The dollar amount on the waiver must equal the retainage being released. The through-date must be the final date of the sub's work on the project. Mismatch on any of those three — wrong form type, wrong dollar, wrong date — and the column doesn't flip.

Lower-tier waivers from sub-subs and suppliers. Mechanics-lien rights persist from the sub's downstream tier even after the prime sub has signed off. A supplier the GC never contracted with can still file a lien against the project for unpaid materials supplied through the sub. Collect final unconditional waivers from sub-subs and material suppliers above a materiality threshold the GC sets per project — commonly any tier with a contract or supply agreement above a stated dollar amount, with the threshold sized to the project's scale. This is the gate most often skipped because it requires the GC to run a parallel collection on parties it has no privity with, and it is the gate whose failure costs the most when a lien lands after release.

Surety consent on bonded subcontracts. When the sub posted a payment or performance bond, the bond's surety typically requires a consent of surety to final payment before the GC releases retainage. Releasing without surety consent can void the bond — the surety's reasoning is that retainage release reduces the leverage the surety would otherwise have to compel cure of any latent defect or unpaid obligation, so consent is the surety's price for keeping the bond in force. The tracker carries a bonded-flag column and a surety-consent-received column; on bonded subs both must be green before release.

Final WH-347 on Davis-Bacon work. Federal Davis-Bacon-covered projects require certified payroll on form WH-347 through the sub's final week on site. A missing or non-compliant final WH-347 exposes the GC, as the prime, to DOL audit risk and potential withholding of contract funds at the federal level. The column flips to satisfied only on receipt of a compliant final form, not on a promise that one is coming.

Current COI at release date. The sub's certificate of insurance must be active on the date retainage is released — not the date the work was completed, not the date the contract was signed, but the release date itself. A lapsed COI at release leaves the GC exposed if a covered claim later traces back to the sub's work. The COI tracker that AP maintains for ongoing project compliance feeds this column directly; the broader workflow for keeping the COI tracker current is covered in the subcontractor COI tracker with ACORD 25 review for GC AP walkthrough.

Warranty acknowledgement. Many subcontracts require the sub to sign a warranty-period acknowledgement at final payment that confirms the warranty start date, the warranty duration, and the sub's obligation to return for warranty work. Releasing without it weakens the GC's leverage on warranty claims six or twelve months later — the sub's defense becomes "I never agreed in writing the warranty had started," and the GC is left arguing implication.

Year-end-spanning 1099-NEC reconciliation. For subs treated as 1099 vendors, releases that cross a calendar-year boundary need the prior-year 1099-NEC totals reconciled before the release flips into the current-year total. A release in the first quarter that should have been recorded against the prior year creates a mismatch with the IRS filing the GC has already submitted, and amending the filing is more expensive than holding the release until the reconciliation closes.

The workpaper at any point in the project's run-up to substantial completion shows, per sub, which gates are satisfied and which are outstanding. The release-target date is computable only when every gate is green; until then, the row stays in pending and AP knows exactly which document or attestation is blocking it.

The tracker structure, and where extraction value lands

The release-schedule workpaper is two tabs joined on project ID. Building it cleanly the first time is faster than retrofitting columns when a project hits substantial completion and the controller realises the trigger rule was never captured.

The project tab carries one row per active project. Columns, in the order they should sit on the sheet:

• project ID
• state
• project type (private / public / federal)
• contract size
• owner name
• prime contract retainage cap (rule and percentage)
• release-trigger rule (substantial-completion-permits-reduction / final-acceptance-only)
• prompt-payment clock (days post-receipt)
• substantial-completion target date
• substantial-completion actual date
• final-acceptance target date
• final-acceptance actual date
• owner-retainage-outstanding amount
• owner-retainage-released-to-date amount

The sub tab carries one row per active subcontract on each project, joined to the project tab by project ID. Columns:

• project ID (foreign key)
• subcontractor name
• subcontract value
• retainage withheld percentage
• retainage withheld to-date
• retainage paid to-date
• retainage outstanding
• substantial-completion-release-permitted flag (per subcontract, set on award)
• substantial-completion-release amount
• substantial-completion-release date
• punch-list-cost holdback amount
• final unconditional lien waiver received
• lower-tier waivers received
• surety consent received (if bonded)
• final WH-347 received (if Davis-Bacon)
• COI active at release date
• warranty acknowledgement signed
• 1099-NEC reconciled (if year-spanning)
• release-target date
• release actual date

The data flows are the part that has to be designed deliberately. The project tab's retainage-cap and release-trigger cells are populated from the state-law reference table maintained per the rule landscape above, refreshed at least annually. The owner-retainage cells are populated from the pay-app history — the cumulative retainage line on each G702 the GC issued to the owner. The sub tab's withheld and paid columns are populated from each sub's pay-app history (the G703 line items if the project runs AIA documents, or the equivalent on a non-AIA cover sheet) and from the release events as they fire. The gating columns are populated from the lien waiver log, the COI tracker, the certified-payroll review packet, and the surety-and-warranty document log.

This is where document-extraction value lands honestly. The pay-app history sits in PDFs that AP teams already process in volume — the G702 retainage line and the matching G703 sub-pay-app line are structured fields on consistently formatted forms, and pulling them into the receivable and payable columns turns a manual roll-forward (read each PDF, transcribe the retainage line, sum across pay apps) into a derived calculation. Final unconditional lien waivers arrive as PDFs on a per-release basis, often with the dollar amount, the through-date, and the form designation in fixed positions on each jurisdiction's statutory form; pulling those three fields into the gating-status flag converts a manual review of every waiver into a derived flag the AP supervisor checks on exception. WH-347 packets, COIs, and warranty acknowledgements are similarly document-shaped inputs to their respective gating columns. The product side of our shop builds tooling to extract pay-app retainage history and final lien waiver status into the release tracker — it doesn't replace the workpaper, it feeds the columns the workpaper depends on.

The rest of the tooling stack sits separately. Most mid-market GCs run the tracker itself in Excel or in their ERP retainage module — Sage 100/300 Construction, Viewpoint Vista, ComputerEase, Foundation, Procore Financials. The extraction value is at the document-to-flag and document-to-cell conversion layer, not at the tracker-host layer. Choosing where the tracker lives is a separate question; our AP automation software buyer's guide for construction companies walks the platform-level comparison for that decision.

Designing each upstream workpaper with a clean lane into a specific gating column is what keeps the release schedule honest as a control surface rather than a parallel system that drifts from the underlying documents.

---

Common errors that expose the GC

Each of these errors is a release event the workpaper would have blocked if the gating columns had been treated as hard gates instead of soft reminders. They surface most often on projects where the controller inherited an in-flight tracker from a predecessor and didn't audit the column definitions before using it to fire releases.

Releasing before lower-tier waivers land. The prime sub's final unconditional waiver was collected and the AP team treated that as the lien-waiver gate satisfied. Weeks after release, a supplier the sub had been buying steel from files a mechanics lien against the project for unpaid materials. The supplier had no privity with the GC and was never on the AP team's collection list. The lower-tier-waivers column on the tracker should have been a hard gate that required collection of final unconditional waivers from sub-subs and material suppliers above the project's materiality threshold; treating the prime sub's waiver as covering the tier below it is the most common version of this failure.

Releasing before surety consent on a bonded subcontract. The sub posted a payment and performance bond at award; AP released retainage at substantial completion without obtaining the consent of surety to final payment. A latent defect in the sub's work surfaces six months later, and when the GC turns to the bond, the surety asserts that the unauthorised release of retainage discharged its obligation. The bonded-flag column and the surety-consent column should have blocked the release; the bonded-flag column is set on award when the bond is recorded and persists through the entire project, so missing the surety-consent gate at release is purely a workflow failure.

Releasing before the final WH-347 on Davis-Bacon work. The project was federally funded and Davis-Bacon-covered; the sub's final week of certified payroll was either missing or non-compliant with the prevailing-wage requirement. The release fired anyway because AP didn't read certified payroll and treated the final WH-347 column as someone else's responsibility. A subsequent DOL audit assesses back wages, interest, and penalties against the GC as the prime, with potential withholding of contract funds at the federal level. The final-WH-347 column should have flipped to satisfied only on receipt and review of a compliant final form, with the certified-payroll review packet feeding the column directly.

Releasing without warranty acknowledgement. The sub signed the final unconditional lien waiver but never signed the warranty-period acknowledgement that would have confirmed the warranty start date and the obligation to return for warranty work. A warranty claim arises six months later; the sub argues the warranty period started months earlier than the GC contends and is now expired, or that the obligation to return was never explicitly accepted. The GC's leverage to compel cure is materially weaker without the signed acknowledgement, and the cost of the warranty work shifts back to the GC. The warranty-acknowledgement column should have gated the release, with no exceptions for subs the GC trusts informally.

Releasing across calendar years without 1099-NEC reconciliation. A year-end-spanning release fires in the first quarter and flips into the current-year 1099-NEC total without reconciliation against the prior-year amount the GC already filed with the IRS. The sub's 1099 is now wrong on the prior-year filing; amending it costs the GC accountant's time, potentially triggers correspondence with the IRS, and creates downstream confusion when the sub's tax preparer reconciles the filed 1099 against the sub's books. The 1099-reconciled column on year-spanning sub rows should have been satisfied — meaning the prior-year total was confirmed and any prior-year-attributable portion of the release was recorded against the correct year — before any cross-year release flipped through the workflow.

The meta-pattern is the same in every case: a release event scheduled by date and reconciled to the gating list afterward, instead of a release event computed only when every gating column was already green. The discipline that prevents the catalogue is to treat the release-target date as a derived value — produced by the tracker once every gate flips — rather than a planned value that AP fires on schedule and chases the gates to catch up. Built that way, the tracker is the control surface; built any other way, it is a record of releases the GC has to defend after the fact.