DOL publishes real figures about the PERM queue, and most of the confusion around processing times comes from reading them as something they aren’t. This guide walks through each number on the data overview, each of the data pages, and each of the calculators, in the order a case meets them.
Start with your own case number
PERM case status takes one PERM case number and shows what DOL holds against it: the status in DOL's own words with a plain-English reading beside it, how many undecided cases were filed ahead of it, and the sponsoring employer's published record. The statuses come from DOL's own case-status search, which PERM Tracker runs in batches against every undecided case twice a day, so a case decided in the last few hours shows at DOL first.
Two limits before you type a number in. Current case numbers look like
G-100-26125-868956, and the shorter 2022 and 2023 form, A-23043-00641,
is looked up too, but its middle block isn't a readable filing date, so no
queue position is drawn from one. And a status alert can only be set on a
case that can still change, because there's nothing to report on a case DOL
has already decided.
The queue position
The headline figure on the overview is the filing month DOL is deciding right now in analyst review, the queue nearly every case sits in. Subtract your filing month from that month and you have your rough distance in queue terms. The tape below it draws the same fact: cleared months solid, waiting months empty, the frontier flagged.
Two cautions. The position is a month, not a promise. DOL clears a month over weeks, ordered alphabetically by employer inside it. And audited cases sit in a separate, slower queue with its own published position.
The average, and why it disagrees with everyone
DOL also publishes an average number of days to a determination. As of August 2026 it said 372, while public tools reported anywhere from 335 to 503. The spread has mechanical causes (different populations, different windows, different data ages), and the methodology page takes them apart one by one. When two numbers disagree, the first question is what each one is measuring, and the second is when its data was published.
How much is still waiting, and how fast DOL is clearing it
DOL's quarterly disclosure files carry a decision date on every row and no pending rows at all, so the size of its own backlog can't be derived from them at any level of effort. The backlog page is built from a per-case scan instead: how many cases are still undecided in each filing month, and which DOL queue each one is standing in.
Decision activity is the other half of the same question. It counts the decisions DOL actually issued, by day and by week, the shape of its working week, and what happened to the cases it decided. It reports what has already happened and forecasts nothing, because the rate the queue moved last month is not a promise about next month.
RFI, audits and appeals
The review-stage page covers the statuses DOL puts in front of an applicant with no explanation attached: RFI issued, pending audit response, supervised recruitment, reconsideration, BALCA. Each one gets what it means, who has to act, the deadline the regulation sets, and the section it comes from. Beside that is a live count of how many cases are standing at each stage, which exists nowhere else because the disclosure files hold no pending cases.
The page publishes no audit rate and no probability that a given case draws an RFI. A rate over a group isn't a forecast for a member of it, and PERM Tracker's own audit measurement currently reads zero, which is a hole in the data rather than a finding: DOL's processing times page publishes a live Audit Review queue, so the true figure plainly isn't zero.
Which calculator answers which question
Before filing. The prevailing wage queue is a count, not a model: how many wage requests sit ahead of the month yours was received. The deadline calculator is the one exact tool in the set: recruitment windows, the quiet period and the ETA-9089 filing window are arithmetic in the regulations, so those dates are computed, not estimated.
While waiting. The decision estimator shows every model its data supports for your filing month, labelled, and lets them disagree. If you want one email when DOL reaches your month, set a queue alert. It fires once and nothing else follows it.
After certification. The I-140 calculator puts USCIS's published processing time next to its own queue arithmetic, and the gap between the two is the useful part. The priority date calculator shows where a date stands against the visa bulletin and which way the cutoff has been moving, backwards more often than people expect.
The whole picture. The green card timeline draws every stage to scale and is honest about the one nobody can number: the wait for a visa number depends on the bulletin, so it’s shown as a question, not a figure.
The whole field: states, wages, sponsors
The disclosure files carry more than timing, and three pages cut the same quarter three ways. The state map places every filing at its worksite state - volume and wages swing hard by state, median days barely move, because DOL works one national queue wherever the case was filed. The wages page gives median offered wages by occupation, which are committed figures from federal filings rather than survey estimates. And the employers page ranks every sponsor in the window, searchable, with each one's approval rate and median days - useful for checking a sponsor's track record, useless for predicting your own case's pace, which only the filing month decides.
Who filed it, and what gets denied
Two more cuts of the same files answer questions the aggregates above can’t. The law firms page ranks every firm filing PERM cases, which is the only public benchmark an attorney has for their own practice against the field, and a way for a beneficiary to see whether the firm on their case has done this before. The denial rates page shows what actually gets denied: by offered wage, by fiscal year, and by the three questions the ETA-9089 asks about layoffs, ownership interest and full-time status. Denials are rare and concentrated, and none of those rates is a probability for a single case.
For caseloads
Every date the deadline calculator produces, the tracker computes automatically per case, with alerts and calendar sync. The public calculators and the tracker run the same regulatory arithmetic from one shared engine, so the two can never disagree.
