By Patrick Turcotte, COO, Docutrax. Last reviewed: 18 September 2026.

Excess liability insurance provides limits above an underlying liability policy. It does not broaden what is covered. For a project owner, the layer only helps if it sits over the right underlying policies, on terms that carry your additional insured status upward, with limits that have not already been eroded elsewhere.

Most published explanations of excess liability are written for the party buying it. This one is written for the party requiring it in a contract and then having to confirm it is real.

What is excess liability insurance?

An excess liability policy is a policy issued to provide limits in excess of an underlying liability policy. That is the whole of its job. It supplies additional limits. It does not expand the scope of coverage beyond the policy underneath it.

The policy underneath is called the underlying insurance, and the specific policies it consists of are listed on a schedule of underlying insurance attached to the excess policy. That schedule sets the attachment point: the amount that has to be paid or established below before the excess layer has any obligation at all.

Two consequences follow immediately, and both matter more to an owner than the limit does.

If a claim is excluded by the underlying policy, a true excess policy does not pick it up. Zero underneath means zero above.

And if the excess policy is scheduled over policies other than the ones you required, the limit on the certificate is sitting above something you never asked for.

How is excess liability different from umbrella coverage?

These get used as synonyms constantly, including on the certificate form itself, and the difference is real.

An umbrella liability policy is designed to protect against catastrophic loss and does three things. It supplies additional limits above the underlying policies. It provides limits when an underlying aggregate has been exhausted. And it may cover some claims the underlying policies do not, in which case the insured funds a self-insured retention first.

A true excess policy does only the first two. It cannot be broader than what sits beneath it.

One thing neither of them does. Umbrella insurers, in Craig Stanovich's phrasing, "do not generally 'drop down' and provide coverage over known uninsured liability exposures." An umbrella covering something the underlying policy excludes is not the same thing as an umbrella covering something the contractor simply never bought. The first happens, subject to the retention. The second generally does not.

Why does follow form wording decide what the layer actually covers?

An excess liability follow form policy is excess insurance subject to all the terms and conditions of the policy beneath it. A stand-alone excess policy provides excess coverage according to its own terms instead.

In practice the line between them is blurrier than those two definitions suggest, and the blur runs one way. Stanovich notes that even where an excess policy expressly promises to follow the terms, conditions, exclusions, limitations and definitions of the underlying insurance, that promise is invariably qualified, and he describes this as usual and customary practice among excess insurers rather than trickery. Gregory Podolak puts it more bluntly: follow-form policies almost always state that the excess policy follows the terms of the underlying policy "except to the extent that the excess policy's terms differ."

Read that carve-out carefully and note which direction it runs. It is what lets the excess policy be narrower than the primary. Stanovich states the traditional distinction directly: excess liability policies do not provide broader coverage than the policies over which they are excess.

The standard form says the same thing about exclusions in so many words. The ISO Commercial Excess Liability Coverage Form, CX 00 01 04 13, provides that "the following exclusions, and any other exclusions added by endorsement, apply to this Coverage Part," and that the exclusions applicable to any "controlling underlying insurance" apply to the excess layer unless superseded by them. So an excess layer can carry an exclusion the CGL beneath it does not carry, and nothing on the certificate will say so. On construction work the ones worth looking for at that layer are exclusions touching residential exposure and action-over claims, because those are the ones that decide whether the layer responds to the loss the owner is actually exposed to.

The other reason wording matters: ISO has published standard umbrella forms since late 2000, but as Stanovich notes, many insurers "still draft and use their own policy forms, or at least draft and use their own endorsement forms." The ISO Commercial Liability Umbrella Coverage Form, CU 00 01, is a real and common starting point. It is not a safe assumption about any particular policy. His conclusion is the operative one: there is no substitute for reviewing the actual terms and conditions of the policy itself.

What does the ACORD 25 show about an excess layer, and what does it not show?

The ACORD 25 gives the excess layer a single row. It is labelled for both, UMBRELLA LIAB and EXCESS LIAB, with checkboxes for which one it is, checkboxes for OCCUR or CLAIMS-MADE, a field for EACH OCCURRENCE, a field for AGGREGATE, and a field for DED or RETENTION.

That is genuinely useful information. Which of the two it is, whether the trigger is occurrence or claims-made, the limits as written, and whether the insured funds a retention before the layer responds are all worth having, and all four are checkable in seconds.

Here is what the row structurally cannot tell you.

It carries no schedule of underlying insurance, so it cannot establish what the layer attaches over. It does not show exclusions, so it cannot establish whether the follow form carve-out has narrowed anything. It shows limits at issuance rather than what remains after prior claims have eroded the aggregate, a point the form itself concedes when it states that limits shown may have been reduced by paid claims. And it does not distinguish a deductible from a self-insured retention even though the field is labelled for both, which matters because under a self-insured retention the insured pays defense and indemnity costs until the retention is reached, while under a deductible the insurer pays and seeks reimbursement afterward.

The form is explicit about its own standing. It states that it is issued as a matter of information only and confers no rights upon the certificate holder, that it does not amend, extend or alter the coverage afforded by the policies listed, and that a statement on the certificate does not confer rights to the certificate holder in lieu of the endorsement. Its cancellation wording says notice "will be delivered in accordance with the policy provisions," which commits the insurer to nothing in your favour.

None of that makes certificate review pointless. A certificate can be the right level of review where the approved requirements and the relationship's inherent exposure do not call for deeper analysis. But an excess layer on a large or high-exposure project is close to the definition of a case where the exposure warrants closer scrutiny, because every question that decides whether the layer works sits in a document the certificate does not contain.

Does your additional insured status carry up to the excess layer?

This is the question owners get wrong most often, and the answer runs opposite to what most people assume.

Additional insured status on the contractor's CGL comes from an endorsement on the CGL, typically a CG 20 10 or CG 20 37 depending on whether ongoing or completed operations are in view. Whether that status reaches the layer above is decided by the excess policy, and the standard forms decide it in the owner's favour with conditions attached. The ISO Commercial Liability Umbrella Coverage Form, CU 00 01 04 13, and the ISO Commercial Excess Liability Coverage Form, CX 00 01 04 13, both provide that any additional insured under the underlying insurance will automatically be an additional insured under the excess layer. No separate umbrella endorsement, no scheduling, no listing.

The conditions sit in the same paragraph of those forms, and both of them narrow what you get. Additional insured coverage provided by the excess layer will not be broader than the coverage the underlying policy provided, so a narrow CGL endorsement stays narrow all the way up. And where additional insured coverage was required by a contract, what the insurer pays on the additional insured's behalf is capped by reference to the amount the contract required, net of amounts payable by the underlying insurance, and it can never exceed the excess policy's own available limits. A contract that names a primary limit and says nothing about the layer above it can therefore cap what you reach there, whatever the umbrella limit says.

None of that is safe as an assumption, because many umbrellas and excess policies are not written on the ISO forms. Podolak's conclusion after working through the wording is that additional insured coverage under excess policies can be more restrictive than under primary policies, which can present exposure considerations for upstream and downstream parties alike. He also notes that where a policy carries no follow form language at all, an additional insured's coverage depends on how the excess policy defines insureds in its own Who Is An Insured section.

The practical version. A certificate showing you as an additional insured, with a $10 million umbrella on the row below, does not tell you which kind of form that second row is, and the difference decides whether you are an insured there and for how much. Confirming it means reading the excess policy alongside the CGL endorsement it points back to.

Why does "primary and noncontributory" often fail at the excess layer?

Requiring primary and noncontributory coverage is standard, correct, and mostly ineffective one layer up unless the excess policy is endorsed to address it. The reason is mechanical rather than anyone acting in bad faith.

Primary and noncontributory on the CGL comes from specific endorsement language, the ISO CG 20 01 or its equivalent. The umbrella or excess policy is a different policy with its own Other Insurance condition, and as Stanovich states, that policy "includes its own Other Insurance condition and will not follow the underlying CGL's 'primary and noncontributory' wording." The assumption that a follow form umbrella inherits the CGL's primary and noncontributory wording is, in his words, flawed, because the umbrella's own Other Insurance condition will usually control.

What that does to the order of payment is worth setting out explicitly, because it is the reverse of what the contract intended.

What the contract intends: the contractor's CGL responds, then the contractor's umbrella, and your own CGL is not touched.

What usually happens: the contractor's CGL responds, then your own CGL, and the contractor's umbrella then responds in conjunction with your own umbrella, prorating the remaining damages between the two umbrella insurers.

ISO's Noncontributory, Other Insurance Condition endorsement does not solve this. It addresses only the right of one insurer to seek contribution from another, and Stanovich's assessment is that CU 24 78 falls short as respects the order of coverage, which is the part an owner cares about. There is, however, a form that addresses it directly. ISO filed Noncontributory and Order of Response, Other Insurance Condition (CU 24 77 12 23), effective 1 December 2023, with a comparable endorsement for the excess form (CX 24 33 12 23). It specifies that the umbrella or excess policy responds before any other insurance available to an additional insured on which that additional insured is a named insured. Two conditions attach: the endorsement schedule has to be completed listing the additional insured, and there has to be a written primary and noncontributory agreement for it to apply.

None of this makes the requirement wrong to include. It makes it something to name at endorsement level and verify at the excess policy rather than assume from the primary.

What happens when the underlying limits have already been used?

Two separate mechanisms live here and they are often collapsed into one.

Erosion. A general aggregate is a shared pot across a policy period. If claims from other projects have already consumed part of it, the limit on your certificate is not the limit available to you. The certificate has no field for this and cannot be made to show it.

Whether that erosion lowers the excess policy's attachment point depends on the excess wording. Stanovich contrasts favourable wording, which recognizes "any reduction or exhaustion of aggregate limits by payment of loss in claims or suits covered by the Underlying Insurance," with narrower wording restricting recognition to losses "to which this policy applies." Under the narrower version, an underlying policy can pay out on something the excess layer would never have covered, the aggregate drops, and the attachment point stays where it was. The difference falls on the insured, and through the insured, on you.

Exhaustion. Exhaustion is not the same as the underlying insurer writing a cheque. Under the wording in common use, primary insurance may be exhausted without full cash payment by the underlying insurer once the obligation to pay damages above the underlying limit is established by final settlement, judgment, or written agreement. Some policies override this with express wording requiring exhaustion through payment in legal currency, which matters a great deal if the underlying insurer settles below its limit. Courts split on the question as well, some permitting exhaustion where the insured settles below the underlying limit and funds the gap itself, others requiring the underlying insurer to actually pay its full limit, so this is one to read in the policy and in the jurisdiction rather than assume.

Insolvency. If the underlying insurer fails, umbrella policies typically state that its bankruptcy or insolvency does not relieve the umbrella insurer of its obligations and that the insurance applies as if the underlying insurance were in full effect. Read the direction of that carefully. The umbrella stays exactly where it is. It does not drop down to cover the failed layer. The gap belongs to the insured.

This is why carrier financial strength on the underlying policy is an excess layer question, not only a primary one.

Whose insurance responds first when you have your own excess tower?

Owners on large projects usually carry their own excess coverage. When a loss reaches into the excess layers, the question of which tower is tapped first has real money attached, and there is no single national answer.

Two competing approaches exist. Horizontal exhaustion holds that the downstream party's excess policy is not triggered unless all applicable primary policies have been exhausted, including the upstream party's own primary CGL. Vertical exhaustion holds that the primary and excess policies purchased by the downstream party must pay before any policy purchased by the upstream party.

Which one governs is jurisdictional. A fifty-state survey by Saxe Doernberger & Vita, current as of September 2024, places California, New York, Illinois and New Jersey in the horizontal camp in the additional insured context, and places Arkansas, Kentucky, Missouri, Texas and Virginia in the vertical camp. Most other states have no decision on the question at all.

Two decisions show what the horizontal rule does to an upstream party in practice. In Kajima Construction Services, Inc. v. St. Paul Fire and Marine Insurance Co., 227 Ill. 2d 102 (2007), the Illinois Supreme Court affirmed summary judgment for the excess insurer where a subcontractor had given the general contractor a certificate naming it as an additional insured with $2 million of general liability and $5 million of umbrella coverage. Because the general contractor also carried its own primary policy, it had to exhaust that policy before reaching the subcontractor's excess coverage. In Bovis Lend Lease LMB, Inc. v. Great American Insurance Co., 53 A.D.3d 140 (1st Dept 2008), New York's First Department reached the same order on an appeal from summary judgment, holding the additional insured coverage afforded by the umbrella policy to be excess to the upstream parties' own primary insurance.

One clarification worth making because the terms collide. California's Supreme Court adopted a vertical exhaustion rule in Montrose Chemical Corp. of California v. Superior Court, 9 Cal. 5th 215, decided 6 April 2020, holding that where a loss is continuous across policy periods, all primary insurance has been exhausted, and the policy language permits it, a policyholder may access coverage under any excess policy once it has exhausted the directly underlying excess policies for the same policy period. That case concerned one policyholder's own tower across many policy years in a continuous-injury claim, and it did not address additional insured coverage. It is a different question from whose tower goes first as between an owner and a contractor, and it should not be read as settling the second.

For an owner, the operational point is narrow and useful. Where your project sits changes the order in which your own insurance gets tapped, and that is worth knowing at contract stage rather than at tender.

Is the excess carrier admitted, or is it surplus lines?

An excess or umbrella layer may be written by a non-admitted carrier in the surplus lines market, which NAIC describes as consisting of insurers covering risks not available in the admitted market.

The difference that matters to a certificate holder is one line long. NAIC states that a consumer protection available in the admitted market "but not available to the surplus lines market, is the protections of a state guaranty fund." If a non-admitted carrier becomes insolvent, there is no guaranty fund behind the policy.

This is not an argument against surplus lines placements. NAIC also notes the insolvency rate of surplus lines insurers is historically low, and for risks the admitted market will not write, the surplus lines market is where the capacity is. It is an argument for two specific checks. Confirm the carrier's AM Best rating rather than assuming it, and know which of your requirements can realistically be met by an admitted carrier before you write a requirement the market cannot fill.

The ACORD 25 does show the insurer name and NAIC number, so this particular check is one the certificate genuinely supports.

How should an owner set excess limit requirements?

Setting a number is the owner's decision and depends on the project. What follows are the factors that change the answer, not a recommendation.

Jurisdiction. Some states impose duties on owners directly rather than derivatively. New York Labor Law section 240(1) reaches "All contractors and owners and their agents, except owners of one and two-family dwellings who contract for but do not direct or control the work," and requires them to furnish scaffolding, hoists, stays, ladders and similar devices in the erection, demolition, repairing, altering, painting, cleaning or pointing of a building or structure. New York's Court of Appeals has held that the duty section 240(1) imposes is nondelegable, that a breach of it leads to absolute liability, and that the statute is aimed only at elevation-related hazards (Ross v. Curtis-Palmer Hydro-Electric Co., 81 N.Y.2d 494 (1993)). An owner exposed under a statute in its own name is in a different position from one exposed only through a contractor's negligence, and required limits usually reflect that.

Whether the aggregate is dedicated to your project. A general aggregate shared across a contractor's whole book is not the same as one that is not. The designated construction project general aggregate limit endorsement, ISO CG 25 03, makes the CGL general aggregate apply separately to each designated project. Note two limits on it. It does not modify the products-completed operations aggregate, and whether the excess aggregate follows on the same per-project basis depends on the excess policy's own wording rather than following automatically.

Completed operations tail. Construction claims arrive years after closeout. A limit adequate at substantial completion is a limit that has to survive the tail, and both the underlying completed operations aggregate and the excess layer above it have to still be in force and still have room.

How the requirement is expressed. A requirement that names a total amount and permits it to be met by any combination of primary and excess is easier to satisfy and easier to verify than one that specifies a structure the contractor's program does not have. It also avoids the outcome where a contractor technically complies with a structure that leaves your additional insured status capped at the primary layer.

Wrap-up programs change all of this, because the coverage structure moves. That deserves its own treatment rather than a paragraph here.

Is excess liability the same as owners and contractors protective liability?

No, and the confusion is worth clearing up because owners searching for one often mean the other.

Owners and contractors protective liability coverage, usually called OCP, is a separate stand-alone policy. The contractor buys it and the owner is the named insured on it. It covers the named insured's liability for bodily injury and property damage caused by the independent contractor's work, and liability arising out of the named insured's general supervision of that work.

The structural difference from everything else in this guide: with OCP you have your own policy, not a position on somebody else's.

Stanovich's comparison of the two approaches is direct in both directions. On the upside, OCP policy limits "are dedicated exclusively to the named insured" and are not shared with other insureds or eroded by claims that have nothing to do with your project. On the downside, the OCP provides substantially less protection than a CGL would, its coverage is confined to vicarious liability and general supervision, it is premises and operations coverage that ends when the work at the site is complete rather than carrying a completed operations tail, and the contractor's umbrella insurer will not usually provide limits to the OCP named insured in excess of the OCP policy. His conclusion is that OCP "is not designed to be a substitute for a CGL; it is most effective when written in conjunction with a CGL."

So the trade is dedicated limits against narrower scope and usually no excess layer above them. Neither is a straight upgrade on the other, which is why the two are usually seen together rather than as alternatives.

What to verify on an excess layer before work starts

A certificate can be the right level of review where the approved requirements and the inherent exposure do not call for deeper analysis. Where scope, jurisdiction, completed-operations exposure, or potential liability warrants closer scrutiny, the certificate may not reveal exclusions or endorsement language that matters at claim time. On an excess layer carrying real limits, most of the questions below sit in the second category.

From the certificate. Whether the layer is presented as umbrella or excess. Occurrence or claims-made trigger, and if claims-made, the retroactive date and reporting requirements. Limits as written and whether the aggregate is separate. Whether a deductible or a self-insured retention applies, and how much. The carrier name and NAIC number, from which the AM Best rating can be looked up. Policy period concurrency with the underlying policies, since a gap of even a few days between them is a gap in the tower.

From the policy. The schedule of underlying insurance, which is the only document that establishes what the layer actually attaches over. Whether it is follow form, stand-alone, or follow form with carve-outs, and what the carve-outs remove. Whether you are an additional insured at this layer, on what terms, and subject to what cap. The Other Insurance condition, which decides the order of response regardless of what the primary's endorsements say, and whether an order of response endorsement has been attached. The maintenance of underlying insurance condition, and what happens if it is not met. The wording governing whether an eroded underlying aggregate lowers the attachment point.

The pattern in that second list is the point. Every question that determines whether the excess layer works is answered in a document the certificate does not contain.

Where excess review fits in a compliance operation

Depth of review should track the exposure, not a fixed rule. Most third-party relationships do not need anyone reading an excess policy. A trade contractor on a large project, in a jurisdiction that puts duties on owners directly, with an eight-figure limit sitting above a CGL you have already endorsed yourself onto, is not most relationships.

That is the case for forensic policy review on the scopes of work where it is proportionate: reading the endorsements and the policies together as one coverage stack rather than checking a certificate row in isolation, because the excess layer's behaviour is determined by documents that never appear on a certificate.

Docutrax runs that work as a managed operation. Licensed P&C professionals and CRIS-certified Account Managers perform the review, against the requirements the client has approved, and the platform makes the record visible and auditable across projects. Where a layer does not match what the contract requires, Docutrax works with the third party and its brokers and agents to resolve it and documents what was found. Decisions about site access, work start, payment, and waivers remain with the client.

Docutrax has served more than 2,500 cumulative client accounts, has completed more than 300,000 forensic policy reviews since 2017, has been in market more than 10 years, and is SOC 2 Type 2.