
20 VA Regulations That Help You Read a Disability Decision
These regulations explain common questions about evidence, service connection, ratings and review options. They are a reading guide, not a list that controls every claim.
Clear guides, original reporting and a closer look at VA claims.
23 articles · All sections

These regulations explain common questions about evidence, service connection, ratings and review options. They are a reading guide, not a list that controls every claim.

Court citations become useful when you understand the question each case answers. This guide covers evidence, examinations, ratings and several important limits.

Our extracted Board records show different outcomes under different service-era labels. The comparison does not isolate the effect of service era or presumptive rules.

The Board's own FOIA data shows appeals are not decided in the order they were filed. I matched 218,614 of those decisions to their text to test the usual explanations and to measure the fast lane the files leave out.

The intelligence and logistics categories differ by 10.2 percentage points in our sample. Small groups, missing labels and repeated decisions limit what that means.

Compare direct, secondary, presumptive, and aggravation theories. Map your actual records to the connection question your claim needs to answer.

Combat veterans grant at 37.2%, non-combat at 30.6%. The 6.6-point gap is real but smaller than the § 1154(b) talking points suggest. What it actually does.

Two high-volume BVA judges show a 40.6-point grant-rate spread. The corrected record and decision counts explain why the comparison needs context.

The Board's grant rate went from 26.9% in 2023 to 30.4% in 2025. Denials up 7 points. Remands down 10. The Board is getting faster but slightly tougher.

Air Force veterans grant at 34.1% at the BVA. National Guard veterans at 21.5%. A 12.6-point spread across services that should follow the same statute.
One email per week with new guides, reports and VA updates. Unsubscribe anytime.