German Pension Refund Processing Times: The Data Behind Our Numbers
Short answer: more than three quarters of the 300 most recent completed refunds managed by Germany Pension Refund reached the client escrow account within three months. Based on GPR-PTS-2026-Q3-v3, 229 of our 300 most recent completed paid refunds (76.3%) reached the client escrow account within 90 days of complete submission. Individual processing times vary. The median was 40.5 days; 185 of 300 (61.7%) reached escrow within eight weeks, and 280 of 300 (93.3%) within six months.
Every figure on this page describes completed refunds in a defined recent cohort. This article documents that cohort — what we measure, which cases count, what was excluded, and what the numbers can and cannot support — so you can judge them the way you would judge any statistic. For the claim process end to end, read our complete guide to claiming a German pension refund; this page covers one question only.
The dataset at a glance
Dataset
Germany Pension Refund processing-time analysis, GPR-PTS-2026-Q3-v3 · calculated 25 August 2026
Cohort
Germany Pension Refund's 300 most recent completed paid refunds — each with a documented complete-submission date and a matched client-escrow value date — ordered by escrow value date. The cohort size stays fixed at 300 between quarterly refreshes.
Excluded
Open/unpaid claims, withdrawals, unsuccessful claims, test records, duplicates and unreliably matched records
Clock
Calendar days from documented complete submission (as recorded by Germany Pension Refund) to the client escrow value date. Per Germany Pension Refund's operating records, the DRV decision was generally issued 1–9 days earlier; decision-letter dates were not used.
Results
Median 40.5 days · 185/300 (61.7%) within 56 days · 229/300 (76.3%) within 90 days · 280/300 (93.3%) within 180 days
Nature
Dated, first-party Germany Pension Refund data among completed refunds · refreshed quarterly · not a new-claim completion probability · aggregates only: case-level records remain confidential
Where the clock starts and stops
The clock starts at documented complete submission — the date Germany Pension Refund recorded the complete application as submitted. Measuring from first contact would include each client's document-gathering phase, which varies with how quickly records can be provided; we therefore start the clock only once the application was recorded as complete and submitted, so the figure describes the post-submission period for applications recorded by Germany Pension Refund as complete. “Complete” here is our recording, not the office's: the responsible pension office can still request additional information later, and our complete guide shows where that happens.
Two things are therefore deliberately not inside this measurement: the preparation phase before submission, and the separate 24-month eligibility waiting period, which runs before a claim can be filed at all. Don't add those to the figures below — they answer a different question.
The clock stops at the client escrow value date. In a managed Germany Pension Refund claim, the refund is paid into the client escrow account operated by our German partner law firm. We measure to the value date on which the money arrived there — a date taken from payment records, because it is the consistently verifiable endpoint. Based on our operating records, the pension office's decision usually precedes that payment — generally by 1–9 days — but decision-letter dates were not used in this calculation: measuring to actual money movement is stricter, and a “days to decision” figure would look faster while telling you less. After the escrow value date, the agreed fee is deducted and the remaining balance is transferred to the client's nominated account — that onward banking step is not part of this measurement.
We count calendar days, not working days. Weekends and holidays inside the span count against the clock, not for it.
Which cases count — and which don't
The cohort is Germany Pension Refund's 300 most recent completed paid refunds: managed claims where both endpoints are verifiably on record — a documented complete-submission date and a matched client-escrow value date — ordered by escrow value date, so “most recent” means the 300 refunds that most recently reached escrow. The cohort size is fixed at 300 between quarterly refreshes; if we ever change it, the change comes with a new dataset version and a stated reason.
Being explicit about what was excluded: open or unpaid claims — cases still in progress with no escrow date — are not in this cohort, and neither are withdrawn claims, unsuccessful claims, test records, duplicates, or records that could not be reliably dated or matched. That makes these figures a factual summary of how recently completed refunds ran — not a prospective probability that a claim submitted today will finish within any given threshold.
Why a fixed-size recent cohort rather than an all-time average? Because processes change: office workloads, routing practice and our own preparation standards are not what they were years ago, and an all-time average would blend all of that history into one blurred number. The fixed rule also reduces discretion in future refreshes — the cohort is the 300 most recent completed refunds, whatever they show.
What this cohort is not. These are managed Germany Pension Refund claims, prepared, checked and routed before submission. They are not a sample of all refund claims in Germany, not self-filed claims, and not a Deutsche Rentenversicherung statistic. The figures are first-party, from our own case and payment records, and nobody independent has audited them — which is exactly why we publish the definitions, the version and the dates that let you interrogate them.
The results
In the 300 completed refunds of dataset GPR-PTS-2026-Q3-v3, counted in calendar days from documented complete submission to the client escrow value date:
-
The median was 40.5 days — about 41 days. A median is a midpoint, not an average or a promise: half the measured completed refunds were faster, and half were slower.
-
185 of 300 (61.7%) reached escrow within eight weeks (56 days).
-
229 of 300 (76.3%) reached escrow within 90 days — the basis for the more-than-three-quarters headline.
-
280 of 300 (93.3%) reached escrow within six months (180 days). The remaining completed refunds took longer.
Read the distribution honestly: it has a tail. A claim that needs a file transfer, an account clarification or a confirmation from a foreign insurance authority can sit in the slower part of that tail — sometimes for reasons visible in advance, sometimes not.
What these numbers cannot tell you
They are retrospective. This is a completed-refunds cohort: it summarizes how finished cases ran. It does not state the probability that a newly submitted claim will finish by a given day — claims still in progress are not in it. We publish it because a defined, dated, versioned retrospective figure still beats an undefined slogan.
They cannot tell you your date. We cannot guarantee a decision or payment date, because the responsible pension office controls processing. A managed service can improve preparation and follow-up; it cannot control the office's queue. The median does not estimate the expected completion time for a newly submitted claim.
They are not a pension-office benchmark. The figures describe managed claims and say nothing about how long any other path takes; they are not a statement by or about Deutsche Rentenversicherung.
They will change. Each quarterly refresh replaces the cohort and the dataset version. When fresh data disagrees with an older published number, the published number changes — in whichever direction the data points.
The pension office's part — and why we don't rank the offices
Germany Pension Refund is a private service. We are not part of or affiliated with Deutsche Rentenversicherung or any German government authority — the pension office decides every claim. “The DRV” is a network (DRV Bund, DRV Knappschaft-Bahn-See and 14 regional carriers), and which office decides an international claim is set by statutory allocation and international liaison rules (§§ 126–128a SGB VI), following your insurance history, citizenship and residence — nobody chooses their office. The dataset does not measure why individual cases took the time they did, but in our operating experience some of the variation in the chart can reflect that structure: files transferring between the account-holding and the deciding office, confirmations requested from foreign insurance authorities, account questions that must be resolved before calculation, and genuine queues in which ongoing pension applications take priority over refunds. None of that is criticism — the offices decide claims carefully and charge no application fee.
We don't publish processing times per office, and the reasons are methodological, not diplomatic. You cannot choose your responsible office, so a league table would inform no decision available to you. Routing concentrates entire countries' claims at particular offices, so per-office samples would range from hundreds of cases to a handful — a ranking built on that would look precise and be misleading. And office workloads shift, so a ranking would start misleading the moment it aged. What actually moves a timeline is whether the claim arrives complete, at the recommended first office, with follow-up handled — the part good preparation can influence.
If your case runs long
Six months matters for two legal reasons. Under § 44 SGB I, statutory interest of 4% per year may become payable once the legal conditions are met; it cannot begin before six calendar months have expired after the complete application reached the responsible German pension carrier — a legal rule, not a Germany Pension Refund benefit or guarantee. And under § 88 SGG an inactivity action may become available after six months without sufficient reason, as a case-dependent last resort that is not filed automatically (in our operating history, none has yet been required). Our full pension refund guide's delay chapter explains the sensible order: complete file first, precise status questions next, escalation only after that.
How we keep these numbers honest
-
Definitions before publication. A changing number appears on our site only once its population, measurement period, definition and source are recorded internally. If a figure lacks that record, it stays unpublished — however good it sounds.
-
A reproducible calculation record. For each dataset version we retain the selected cases, ordering rule, endpoints, exclusion log, threshold numerators and formulas internally, so every published figure can be recomputed. The public sees aggregates; the record makes them checkable. Case-level records stay confidential.
-
A fixed cohort rule. The cohort is the 300 most recent completed refunds — fixed between quarterly refreshes. Any change in cohort size would require a new dataset version and a stated reason, making the change visible and documented.
-
Versioned, dated releases. Each quarterly recalculation gets a new dataset identifier (this page: GPR-PTS-2026-Q3-v3) and a visible calculation date.
-
Superseded numbers are retired, not defended. When a fresh calculation produces different results than an older analysis, the older claim comes down — including claims that sounded better. The figures on this page replaced earlier published ones for exactly that reason.
-
Aggregate results only, honest labels. We publish cohort-level figures, never client-identifiable details — and first-party results are labeled as exactly that, never as independent verification or a pension-office statistic.
Frequently asked questions
Is the 40.5-day median a promise for my case?
No. The median is the midpoint of the measured completed refunds: half were faster, half were slower — some considerably. It does not estimate the expected completion time for a newly submitted claim, and we cannot guarantee a decision or payment date, because the responsible pension office controls processing.
Does the 24-month waiting period count as processing time?
No. The waiting period is an eligibility condition that runs before a claim can be submitted at all. The figures on this page start at documented complete submission — don't add the two together.
Why don't you publish processing times per pension office?
Because the result would mislead rather than inform: you cannot choose your responsible office, per-office sample sizes are necessarily very uneven, and office workloads shift over time. The reasoning is in the section above.
Do these figures include the transfer from escrow to my account?
No. The measurement ends at the value date on which the refund reached the client escrow account. After that, the agreed fee is deducted and the remaining balance is transferred to the account nominated for the payout — a separate banking step whose duration depends on destination, currency and compliance checks.
Were these numbers independently verified?
No — they are first-party figures from our own case and payment records, and we label them that way. That is why we publish the cohort definition, the exclusions, the endpoints, the version and the calculation date: so the claim is specific enough to be questioned, rather than vague enough to be safe.
Can I access the underlying dataset?
No. The underlying records are clients’ case and payment data, and they remain confidential. What we publish is designed to be checkable without them: the aggregates, the full cohort definition and exclusions, the measurement endpoints, and a stable version identifier for every quarterly calculation — with a reproducible calculation record retained internally for each version.
Next step: check your eligibility with the free checker and estimate your refund with the calculator. If your claim is already filed and overdue, the delay chapter of our complete guideshows the status questions worth asking first.
Ready to claim now? →