A remand sends an issue back for more work before it can be decided. When the same issue returns to the Board and is sent back again, the useful question is what remains unresolved.
That might be an unanswered medical question, an incomplete search for records or a failure to carry out an earlier instruction. The number of remands does not answer that question by itself. It also does not tell you how the claim will end.
I checked Claim Raven's extracted Board-decision data because the earlier draft of this article implied that repeated remands steadily increased the chance of a grant. The numbers do not support that conclusion.
What the reviewed records show
Our September 16, 2026 snapshot contained 101,518 condition-level analysis records associated with 49,876 distinct Board citation numbers. A decision may address several conditions, so these records are not independent decisions or individual veterans.
The prior-remand field was populated in 64,108 records. The table below groups those records by the extracted count. “Favorable” includes records labeled granted, partial grant or mixed. Its denominator includes granted, partial grant, mixed, denied and remanded outcomes.
| Earlier remands recorded | Condition records | Favorable records | Favorable share |
|---|---|---|---|
| 0 | 47,719 | 17,275 | 36.2% |
| 1 | 10,185 | 3,267 | 32.1% |
| 2 | 3,606 | 1,094 | 30.3% |
| 3 | 1,386 | 434 | 31.3% |
| 4 | 818 | 228 | 27.9% |
| 5 | 229 | 72 | 31.4% |
| 6 or more | 165 | 61 | 37.0% |
The pattern rises and falls. The highest-remand group is also small: just 165 condition records. Another 37,410 records had no value in the prior-remand field; those cannot be treated as having zero remands.
These are groups observed in a database snapshot. We did not follow one cohort through its first, second and third remands. A percentage here is not a veteran's probability of eventual success, and a remanded outcome is unfinished rather than a final loss.
Read the instructions at the end of the decision
A remand order usually identifies both the missing work and the question it is intended to answer. Start there before gathering more material.
Suppose the Board asks an examiner to address whether a service-connected condition aggravated a second condition. An opinion that discusses only whether the first condition caused the second may leave the requested question unanswered. The wording of the actual remand and the explanation in the new opinion matter.
A simple comparison can help:
| Remand instruction | What to look for afterward |
|---|---|
| Obtain identified treatment records | The records, or documentation of the required search and any inability to obtain them |
| Address a particular medical history | A discussion of that history and why it affects the conclusion |
| Consider causation and aggravation | An explanation that responds to each question the Board asked |
| Clarify an examination finding | A new finding or a reasoned explanation of why clarification cannot be provided |
The point is to identify a concrete gap. A disagreement with the final opinion is not, by itself, proof that the remand was ignored.
One long appeal illustrates the distinction
Board decision 25007303 describes a cardiovascular appeal that returned repeatedly from the Court and the Board. The final decision discussed addendum medical opinions, earlier diagnostic findings and attempts to obtain missing records. The Board ultimately denied the claim in May 2025.
That case shows why another remand should not be described as an inevitable step toward approval. Further development may supply evidence supporting a claim, clarify why the evidence does not support it, or reveal that another question still needs an answer.
It also shows why counting procedural events requires care. A Court remand to the Board and a Board remand to the regional office are different events. A database field should not silently treat every return of a file as the same thing.
The review system affects what happens next
Legacy appeals and the modernized review system have different procedures. In the modernized system, the Board generally remands pre-decision duty-to-assist errors and certain other errors that meet the governing rule. A remand is not an unrestricted invitation to restart every part of the claim.
VA explains its current pathways in decision reviews and appeals. For the next steps after a return to the regional office, see our guide to what happens after a BVA remand. Read the notice issued in your own case before choosing a response or assuming a deadline.
An accredited representative can compare the order with the development VA completed. Bring the remand, the new examination or opinion, and the records specifically identified in the order. That is a more focused starting point than asking how many remands are “too many.”
What this analysis can and cannot establish
The table is a descriptive count of Claim Raven's extracted condition records, not an official VA processing or outcomes report. Automated extraction can misread procedural history. Multiple records can come from one decision, and differences between groups can reflect the types of claims and cases selected for analysis.
We checked the aggregation and reviewed public decisions for context. We did not validate every extracted remand count against the complete procedural record. For that reason, the table should be used to understand the limits of this dataset, not to estimate a filing strategy, a waiting time or personal odds.
The actionable question remains the one in your own file: what did the Board direct VA to do, and does the record show that work was completed?
What Happens After a BVA Remand? Tasks and Next Decisions
BVA Case Breakdown: When a C&P Exam Is Inadequate, the Board Sends It Back
One TERA Memo, Seven Reopened Claims, and a $1,678 Rating: Inside BVA 24018149