In Claim Raven's analysis of 63,070 BVA decisions where combat status was classifiable, combat-indicated veterans were granted at 37.2%. Non-combat veterans were granted at 30.6%. That's a 6.6-percentage-point gap.

I expected the gap to be bigger. The way 38 U.S.C. § 1154(b) gets talked about on veteran forums, you'd think combat status was a kind of skeleton key, that proving combat exposure would unlock service connection on conditions where the medical record is thin or the timeline is messy. The data doesn't support that read. The combat presumption helps. It helps less than the talking points suggest, and it helps in specific places more than others. This post walks through what 6.6 points actually means, where in the data the combat effect lives, and where it disappears. I want to be honest that some of what you've been told about combat status is over-promised.

TL;DR

  • Combat-indicated veterans grant at 37.2% (4,627 of 12,426 cases). Non-combat veterans grant at 30.6% (15,508 of 50,644). A 6.6-point gap on 63,070 classifiable cases.
  • The gap is smaller than veterans expect because 38 U.S.C. § 1154(b) only relaxes the evidentiary standard for the in-service event prong. It doesn't help with current diagnosis or with nexus, which is where most BVA claims actually fail.
  • When nexus is strong, combat status adds almost nothing. Combat + strong nexus is 91.5%; non-combat + strong nexus is 90.8%. The nexus opinion is doing the work.
  • When nexus is missing, combat status barely moves the needle. Combat + missing nexus is 2.6%; non-combat + missing nexus is 1.3%. Doubled rate off a tiny base.
  • Combat helps most when it stacks with a presumptive connection. Combat + presumptive = 52.7%. That's where the lever has the most leverage.
  • "Combat indicator" in our data is noisier than you'd think. Many veterans with documented hostile fire don't get tagged as combat in the BVA decision text because the case didn't turn on the combat presumption.
What evidence wins a VA claimFree

Strong medical opinions won 89.7% of the time in our analysis. Weak ones won just 3.7%.

  • Know what makes a medical opinion harder to dismiss.
  • Use your own statement where it can carry real weight.
  • Catch common C&P exam problems before they hurt your claim.
Get the free guide

The 6.6-Point Gap

The headline numbers

Let me start with the topline number. Across 63,070 BVA decisions where I could classify whether the veteran had a combat indicator in the case record:

  • Combat = true: 12,426 cases. 4,627 granted. 37.2% grant rate.
  • Combat = false: 50,644 cases. 15,508 granted. 30.6% grant rate.

The gap is 6.6 percentage points. The combat-indicated grant rate is 21.6% higher than the non-combat grant rate in relative terms.

That's real. It's not noise. A 6.6-point gap on 63,070 cases is well outside the margin of statistical chance.

It's also smaller than what I expected when I ran the query the first time. I have read enough forum posts and seen enough YouTube videos to know that combat status gets discussed as a major advantage in the claims process. The 1154(b) presumption gets framed as a powerful tool for combat veterans, and there's a strong implication that combat exposure makes claims easier across the board.

Six and a half percentage points is not nothing. It also isn't the night-and-day difference the talking points suggest. A combat-indicated veteran at the BVA is granted 37.2% of the time. A non-combat veteran is granted 30.6% of the time. Both of those numbers are well below 50%. The combat presumption raises the floor; it doesn't change the ceiling.

The interesting question is why the gap isn't bigger. The answer is in the statute itself.

Why Combat Status Helps at All: § 1154(b)

Combat-related VA claims get a specific statutory benefit under 38 U.S.C. § 1154(b). The text is short. The actual scope of the presumption is narrow. Both matter.

Here's what the statute does. For any veteran who engaged in combat with the enemy during a period of war (or in any other period or location covered by the statute), VA accepts "satisfactory lay or other evidence" of an injury or disease incurred during combat as sufficient to establish service incurrence, even if there's no official military record of it. The standard is "consistent with the circumstances, conditions, or hardships of such service."

In plain English: if you say a thing happened to you in combat, and the thing is consistent with combat circumstances, VA has to accept that the thing happened. Even if your service treatment records are silent. Even if there's no buddy statement. Even if your DD-214 doesn't directly corroborate the incident.

That's a significant evidentiary lift. It removes one of the hardest hurdles in many claims, proving an in-service event from records that were poorly kept, lost, or never documented in the first place.

Here's what the statute does not do. § 1154(b) does not change the standard for proving a current disability. It does not change the standard for proving a nexus between the in-service event and the current condition. It does not extend the benefit-of-the-doubt rule beyond its normal application under 38 U.S.C. § 5107(b). It does not lower the rating threshold for any specific condition.

In Shedden v. Principi terms, the three-prong test for direct service connection requires (1) a current disability, (2) an in-service event or injury, and (3) a nexus between them. § 1154(b) relaxes the standard on prong 2. Prongs 1 and 3 still have to be met with normal evidence.

That's the structural reason the 6.6-point gap isn't bigger.

The "One Prong of Three" Limit

This is the framing I want you to take from this post if you take only one thing.

When a BVA claim fails, what does it fail on? In our 41,447-case denial pool from the broader dataset, "in service event missing" was the primary denial reason in only 2.5% of classified denials. Six hundred and thirteen cases. That's the slice of denials where § 1154(b) would have been most directly useful, cases where the in-service event was the missing piece.

Six hundred and thirteen cases out of more than forty thousand. About 1.5% of all denials in the dataset, or 2.5% of denials we could classify by primary reason.

Compare that to the denial reasons that dominate. Severity insufficient is the #1 reason at 29.1%. Nexus gap is #2 at 22.9%. Service connection missing is #3 at 17.0%. Diagnosis missing is #4 at 16.3%. The big four together account for 85.3% of classified denials.

§ 1154(b) does not address any of those four. It does not help establish current severity. It does not bridge nexus. It does not establish service connection on its own (only the in-service event). It does not provide a diagnosis.

If most BVA claims fail because of nexus, severity, current disability, or broader service connection issues, and the combat presumption fixes none of those things, the math tells you why the combat-vs-non-combat gap is only 6.6 points. The presumption fixes a small slice of denial reasons, in a small slice of cases.

Combat × Nexus Quality: Where the Gap Lives and Where It Doesn't

The cross-tab that proves what the presumption does

The cleanest way to see the structural limit of § 1154(b) is to cross-tab combat status with nexus quality. If the combat presumption mattered for nexus, the combat-vs-non-combat gap should be visible at every nexus quality level. If it mattered only for the in-service event prong, the gap should disappear when nexus is strong and stay limited when nexus is missing.

The data is consistent with the second story.

Strong nexus quality:

  • Combat = true: 1,795 cases, 91.5% grant.
  • Combat = false: 5,728 cases, 90.8% grant.
  • Gap: 0.7 points. Functionally zero.

Weak nexus quality:

  • Combat = true: 2,272 cases, 4.5% grant.
  • Combat = false: 9,938 cases, 2.8% grant.
  • Gap: 1.7 points. The grant rate is 60% higher in relative terms, but the absolute base is so small that the lift is marginal.

Missing nexus quality:

  • Combat = true: 1,421 cases, 2.6% grant.
  • Combat = false: 6,518 cases, 1.3% grant.
  • Gap: 1.3 points. Doubled rate off a tiny base. In practical terms, combat status moves you from "almost certain denial" to "almost certain denial."

The pattern is unmistakable. When the nexus opinion is strong, the medical evidence is doing the work and combat status adds nothing on top. When the nexus opinion is weak or missing, combat status helps a little but cannot rescue the case. The combat presumption simply cannot do the work that a nexus letter does.

The 6.6-point topline gap, in other words, isn't evenly distributed across the nexus-quality spectrum. It's coming from a different place in the data. That place is the connection-type cross-tab.

Where Combat Helps Most: Direct Connection (47.2% vs 40.3%)

The clearest place to see the combat presumption doing work is in the cross-tab with connection type.

VA recognizes a few different theories of service connection. Direct connection, the condition started in service. Presumptive, the condition is presumed connected based on service in a specified location or time frame (Agent Orange, burn pit, Gulf War, Camp Lejeune). Secondary, the condition was caused or aggravated by another already-service-connected condition. Aggravation, a pre-existing condition was made worse by service.

Here's combat status by connection type:

Direct connection:

  • Combat = true: 5,669 cases, 47.2% grant.
  • Combat = false: 22,609 cases, 40.3% grant.
  • Gap: 6.9 points.

Presumptive connection:

  • Combat = true: 1,873 cases, 52.7% grant.
  • Combat = false: 5,109 cases, 47.2% grant.
  • Gap: 5.5 points.

Secondary connection:

  • Combat = true: 2,177 cases, 39.5% grant.
  • Combat = false: 9,812 cases, 35.2% grant.
  • Gap: 4.3 points.

The 6.9-point lift on direct connection is the biggest single number in the combat data, and it's the cleanest evidence that § 1154(b) does what the statute says it does. Direct connection cases are exactly where the in-service event prong matters most. If a combat veteran's lay statement establishes the in-service event without a corroborating record, that veteran is more likely to win a direct connection claim than a non-combat veteran trying to establish the same event with stronger evidentiary requirements.

47.2% versus 40.3%. That's the combat presumption doing its job, quietly, on a specific slice of the docket, in cases where the in-service event question actually matters.

Presumptive connection adds 5.5 points. That's a stacking effect, combat status helps on the in-service-event prong, and a presumptive theory removes the nexus requirement entirely if the condition is on the presumptive list. The two together produce the highest grant rate in the connection-type breakdown.

Secondary connection adds 4.3 points. That's the smallest combat effect of the three, because secondary claims are about the nexus between an already-service-connected condition and a new condition. § 1154(b) doesn't help with the nexus question, and it doesn't apply to the link between two civilian-era conditions even when the first one is service-connected.

The combat presumption is a tool that fits one job. It does that job well. It doesn't do the other jobs that veterans sometimes expect it to.

The Presumptive Stacking Effect

I want to spend a moment on the combat-plus-presumptive intersection because it's where the combat presumption produces its highest grant rates in the data.

Combat + presumptive = 52.7%. Non-combat + presumptive = 47.2%. Both are well above the overall average grant rate of 30.6%.

Why this combination works so well is structural. A presumptive condition under 38 U.S.C. § 1116 (Agent Orange), § 1117 (Gulf War undiagnosed illness), § 1119 (burn pit / PACT Act), or various other presumptive statutes means VA accepts the nexus between qualifying service and a listed condition as established by law. The veteran doesn't have to prove a medical link. They have to prove the qualifying service and the qualifying condition.

When you stack combat status on top of a presumptive theory, you've made two of the three Shedden prongs nearly automatic. The in-service event is presumed (combat). The nexus is presumed (statute). The only thing left to prove is the current diagnosis at a compensable level.

This is why so many burn pit and Agent Orange claims grant at high rates. The structural lift from stacking presumptions removes most of the friction that drives BVA denials in non-presumptive cases. Combat status compounds that lift, but the presumption itself is doing the heavier work.

For veterans who served in qualifying conditions for presumptive service connection, the lesson is to file presumptively first. The grant rate on presumptive claims is high. The lift from being combat-indicated within a presumptive claim is meaningful but second-order. The first-order move is to establish you fit the presumptive criteria.

What Most Veterans Get Wrong About § 1154(b)

I've read enough forum posts to know what the common misconceptions are. I want to name them.

"Combat status means VA has to accept my claim." No. It means VA has to accept the in-service event prong with lay evidence alone if the event is consistent with combat circumstances. It doesn't mean VA has to grant the claim. Current diagnosis and nexus still have to be established.

"My CIB / CAR / CMB proves combat for VA purposes." Combat badges are powerful evidence, but they're not the only way to establish combat for § 1154(b) purposes, and they're not automatically sufficient on their own. The statute requires the veteran to have "engaged in combat with the enemy," which VA reads as actual participation in events of combat, not just service in a combat zone. The badges help. They don't always close the question.

"If I'm combat-indicated, I don't need a nexus letter." You absolutely still need one. Combat status helps with prong 2 of Shedden. Prongs 1 (current disability) and 3 (nexus) still require their own evidence. The nexus letter is the load-bearing piece in most claims, and combat status doesn't substitute for it.

"If I have a Purple Heart, I can claim anything as combat-related." No. The Purple Heart establishes that the veteran was wounded in action, which establishes both combat and a specific service-incurred injury. That's powerful for the specific injury underlying the Purple Heart and conditions plausibly secondary to it. It does not establish combat-related origin for arbitrary other conditions you might claim later.

"§ 1154(b) gives me the benefit of the doubt." It does, but only for the in-service event prong. The general benefit-of-the-doubt rule under § 5107(b) and Lynch v. McDonough applies to your whole claim already, combat or not. The § 1154(b) presumption is specifically about not requiring documentary corroboration for combat-related events.

The combat presumption is a specific tool with specific scope. It's not a magic wand. Veterans who treat it as a magic wand end up disappointed when nexus failures sink a claim they thought combat status would carry.

Where It Helps Least: Missing Nexus

I want to land the harder version of the combat-status story.

If your claim has a missing nexus opinion, combat status will not save you. Combat + missing nexus is 2.6%. Non-combat + missing nexus is 1.3%. Yes, the rate is doubled. The base is so small that the doubled rate is still effectively a denial.

What this means in practice is that combat-indicated veterans who file claims without a nexus letter, relying on the combat presumption alone, almost universally lose at the BVA. The presumption establishes the in-service event. It does not establish the medical connection between that event and a current condition five, ten, twenty, fifty years later.

This is the trap I see most often in combat-veteran claims that get to the Board and lose. The veteran has a strong combat record. The in-service event is well-documented or established by the combat presumption. The current diagnosis is clear. The case looks like a slam-dunk on the veteran's side. But there's no nexus opinion explaining how the in-service event caused the current condition, and the C&P examiner, operating from the medical evidence rather than the combat record, provides an opinion against service connection.

The Board, applying the law correctly, denies. Combat status earned the veteran the in-service event presumption. The nexus prong failed. Shedden requires all three. The case loses.

The lesson is that combat status is a foundation, not a structure. If you're a combat veteran filing a claim, you still need to build the rest of the file. The combat presumption gives you a stronger foundation than a non-combat veteran has. It does not relieve you of the obligation to obtain a nexus opinion that bridges service to your current condition.

The Combat Indicator Is Noisier Than You Think

I want to flag a data-quality issue that probably understates the combat effect modestly.

Our "combat indicator" field is extracted from BVA decision text. The field is set to true when the Board's decision references combat exposure, the combat presumption under § 1154(b), or a combat-related context that the judge engaged with in the decision. It's set to false when the decision text doesn't reference any of those things.

That's not the same thing as "the veteran was in combat." It's "the BVA decision discussed combat as a relevant factor."

There are veterans in the dataset who served in support MOSs in combat zones, were exposed to direct hostile fire on convoys or in incidents at forward operating bases, and qualify for § 1154(b) but whose BVA decision didn't focus on the combat angle. Maybe their claim was a non-combat-related condition like sleep apnea, where the combat presumption wasn't relevant. Maybe the decision turned on the C&P exam adequacy and never reached the combat question. Those veterans are in the dataset with combat = false even though they would qualify for combat presumption under the statute.

I don't have a clean way to estimate how many veterans this affects. Probably a meaningful share, especially for post-9/11 era veterans where forward operating base attacks and IED exposure were widespread and where combat MOS designations are less predictive of actual combat exposure than they were in earlier wars.

The practical effect is that my "non-combat" group probably includes some veterans who could have invoked § 1154(b) but didn't. That makes the 30.6% grant rate for the non-combat group slightly higher than it would be if I could cleanly separate true non-combat veterans from misclassified combat veterans. Which means the 6.6-point gap I'm reporting is probably a slight underestimate of the true combat effect.

I want to be honest about that direction. The 6.6 number is conservative. The true gap is probably a bit larger. It's not 30 points. It's probably 7 or 8.

That's still a smaller gap than the talking points suggest.

What the Data Can't Tell Me

A few things this post cannot address.

I cannot tell you whether combat veterans have stronger evidence on average for reasons independent of § 1154(b). Combat veterans may be more likely to have CIB / CAR / CMB documentation, more likely to have unit history corroboration, more likely to have buddy statements available, more likely to have already-service-connected combat-related conditions that serve as anchors for secondary claims. Some of the 6.6-point gap is probably this kind of structural evidence advantage, not the § 1154(b) presumption itself.

I cannot tell you what your specific combat exposure will mean for your specific claim. The combat indicator in the dataset is binary. Real combat exposure is on a spectrum, and the evidentiary value of any specific exposure depends on how it's documented and how it relates to the claimed condition.

I cannot tell you that filing a claim as combat-related will make VA process it faster or differently at the regional office level. The dataset is BVA-level only. Selection bias at the appeal level may shift the apparent combat effect, combat veterans may appeal at different rates than non-combat veterans, and the cases that get appealed may not be representative of all claims.

I cannot tell you whether the Bronze Star or other awards that don't carry an automatic combat credential are sufficient on their own to establish combat for § 1154(b). The statute requires actual engagement with the enemy, not just service in a combat zone or non-combat valor awards. Whether a specific award supports combat status is a fact-specific question I'm not able to answer from this data.

I cannot tell you whether VA's interpretation of § 1154(b) is consistent across regional offices. Some ROs apply the combat presumption liberally. Others apply it narrowly. The data at the BVA level reflects the Board's interpretation, not the full range of regional variation.

What This Means If You Are Combat-Indicated

If your service record establishes combat exposure, you have a real but bounded advantage in the claims process. Specifically:

You can establish in-service events from your lay statement alone if those events are consistent with combat circumstances, without requiring corroborating service treatment records. This is the core § 1154(b) benefit.

Your grant rate at the BVA, on average, is 37.2% versus 30.6% for non-combat veterans. A 6.6-point advantage. Meaningful, but not decisive.

Your biggest single advantage is on direct connection claims where the in-service event question matters, 47.2% versus 40.3%, a 6.9-point lift.

Your stacking play, if available, is combat + presumptive service connection. That combination grants at 52.7%, the highest grant rate in the data for combat veterans.

Your obligation, despite combat status, is still to obtain a nexus opinion for direct connection claims. Combat status does not substitute for the nexus prong. Veterans who file without a nexus letter, relying on combat status alone, lose almost universally at the Board.

Your tactical move, if you're filing a claim, is to invoke § 1154(b) explicitly where the in-service event is contested or undocumented. Your lay statement should reference the combat circumstances under which the event occurred. Your representative should cite the statute in any brief.

What This Means If You Aren't Combat-Indicated

If your service record does not establish combat exposure, you don't have access to § 1154(b). You're still in the same evidentiary universe as combat veterans on prongs 1 and 3 of Shedden, current disability and nexus, which is where most claims actually live or die anyway.

Your grant rate at the BVA, on average, is 30.6% versus 37.2% for combat veterans. You're 6.6 points behind. That's a real gap, but it's not insurmountable.

Your structural levers, the ones available to all veterans regardless of combat status, are the same ones that drive most outcomes:

  • Get a nexus opinion that explains the medical link between service and your current condition. Strong nexus produces a 89.5% grant rate across the dataset. That's higher than the combat-veteran-with-strong-nexus number (91.5%) is above the non-combat-with-strong-nexus number (90.8%). The nexus opinion is doing more work than combat status for almost everyone.
  • Write specific, dated lay statements. The Board quotes lay statements in 49% of granted cases. Your lay statement is admissible, valuable evidence, and you don't need combat status to make it credible.
  • File presumptively where applicable. PACT Act, Agent Orange, Gulf War undiagnosed illness, Camp Lejeune, these presumptions are available to all qualifying veterans, combat or not, and they grant at higher rates than non-presumptive claims.
  • Don't assume the non-combat designation is destiny. The 30.6% non-combat grant rate is the average across all conditions and connection types. With a strong nexus, the grant rate climbs to over 90% regardless of combat status. The lift you get from combat status is real but smaller than the lift you get from a well-developed claim.

The 6.6-point gap is the price of not having § 1154(b). It's not an insurmountable disadvantage. The variables you control, evidence development, nexus opinion, presumptive theory selection, lay statement specificity, all matter more than the combat designation.

Bottom Line

Combat-indicated BVA appellants grant at 37.2%. Non-combat appellants grant at 30.6%. The gap is real, statistically meaningful on 63,070 cases, but smaller than the way 38 U.S.C. § 1154(b) gets framed in most veteran conversations. The reason the gap isn't bigger is that the combat presumption only relaxes the evidentiary standard for the in-service event prong of Shedden, and most BVA claims fail on nexus, severity, current diagnosis, or broader service-connection issues that the combat presumption doesn't address. The cleanest place to see the combat lever doing work is on direct connection claims, where the gap is 6.9 points. The cleanest place to see the combat lever failing to do work is on missing-nexus cases, where the gap is 1.3 points off a base of 1.3%. The compounding play is combat + presumptive, at 52.7%. The combat presumption is a foundation, not a structure. If you have it, use it correctly. If you don't, the variables that actually drive most outcomes are still available to you.

Methodology and Limitations

  • Data source: Claim Raven's analysis of 101,518 condition records drawn from 49,876 Board decisions from the production case_analysis dataset. The combat-indicator field was classifiable in 63,070 of those records. The remaining records either lacked a combat reference in the decision text or used language that didn't allow clean classification.
  • Sample size: 12,426 cases tagged combat = true. 50,644 cases tagged combat = false. Cross-tabs with nexus quality and connection type use subsamples that range from 1,421 cases (combat + missing nexus) to 22,609 cases (non-combat + direct).
  • Classification approach: The combat indicator is set when the BVA decision references combat exposure, the § 1154(b) presumption, or a combat-related context the judge engaged with substantively. It is not set when the decision discusses non-combat military service, support-role service in a combat zone where combat wasn't the focus, or where combat may have occurred but wasn't relevant to the decision text. This is a classification from decision language, not a service-record-derived label.
  • Limitations:
  • The combat indicator likely understates the true rate of combat-eligible veterans in the dataset. Veterans who served in combat conditions but whose BVA decision didn't focus on the combat angle are classified as non-combat. This biases the non-combat group's grant rate slightly upward and makes the 6.6-point gap a conservative estimate of the true effect.
  • These are BVA-level cases only. Veterans whose claims were granted at the regional office, who never had to appeal, are not in this dataset. Combat veterans may grant at higher rates at the RO than non-combat veterans, which would not show up at the BVA level.
  • Selection bias from who appeals to the BVA may differ between combat and non-combat veterans. The 6.6-point gap is among appellants, not among all claimants.
  • The connection-type breakdowns assume the BVA decision correctly classified the theory of service connection. In cases where multiple theories were argued, the classification reflects the primary theory the Board engaged with, not the universe of theories the veteran raised.
  • I cannot fully separate the independent effect of § 1154(b) from structural evidence advantages that combat veterans may have for reasons unrelated to the statute (e.g., better service-record documentation, more available buddy statements, anchor conditions for secondary claims).
  • The "missing nexus" subgroup has a grant rate so close to zero that the doubled rate (1.3% to 2.6%) is sensitive to small numerator changes. The gap in this category should be read as "combat barely moves the needle on missing-nexus claims," not as "combat doubles the rate" in any practical sense.
  • Aggregate patterns are not predictions for any individual claim. Your combat status, your specific service record, and the documentary evidence in your file all matter more than any aggregate average.
  • § 1154(b) interpretation can vary across regional offices and across BVA judges. The aggregate combat effect in the data reflects average application, not the full range of variance in how the statute gets applied case-by-case.

Disclaimer

I'm not accredited by VA, not a lawyer, not a VSO. This is data analysis, not claim advice. These are patterns from cases that made it to the BVA, they don't predict individual outcomes. If you need help with your claim, work with an accredited representative.

Where to go next

-Landon Founder, Claim Raven | U.S. Army Veteran