Skip to the email field

OSHA site-specific targeting

Am I on OSHA's inspection list?

You filed your 300A by March 2 and cannot change it. This is where that rate lands inside your NAICS code.

Your own rate, and where it sits in this ordering. If your establishment has no row in the file, that is an answer too.

Plastics filers, by rateSample

lowest rate filedhighest rate filed

middle of the fileDART 6.96

One filed DART rate, plotted against the plastics filers in the file

Every plastics establishment in the file, lowest rate to highest, with one rate marked.

OSHA’s published CY2025 file. The rate is a sample.

383,283 rowsOSHA’s published Form 300A file for CY2025
CPL 02-01-067the targeting directive, effective 2025-07-17
Two of four criteriacomputable outside OSHA, and the board says which
01the criteria board

Two criteria compute; two need OSHA’s cutoffs.

CriterionComputable outside OSHAYour readout
1 · Elevated DART rateNo; unpublished, and different for manufacturing.Your rate and percentile. The cutoff stays unknown.
2 · Upward-trending rateYes: rate against the national average, and its trend.Both halves, from the filings.
3 · Did not file the required Form 300AYes: in the published file or not.Whether yours appears at all.
4 · Low DART rate, sampled to verify the dataNo. That cutoff is unpublished too.Nothing.

CPL 02-01-067 took effect 2025-07-17 and reached us only as a PDF, so these criteria are paraphrased. The cycle in force reads CY2021 to CY2023 filings; the number you filed this March feeds a later list.

02the trigger

March 2, every year

The day after the filing deadline, your number is fixed for another twelve months.

OSHA never tells you where it puts you. The first notice is an inspector at reception, or a prequalification form asking for a TRIR you know is bad.

03how it works

The file gives the rate and the percentile.

  1. 01

    OSHA’s own file, opened

    84.6 MB, 383,283 rows, 106,225 employer identification numbers, no key required. The directive’s scope (non-construction, 20-249 employees) leaves 211,349 rows with usable hours.

    OSHA ITA Form 300A summary data, CY2025 through 2026-03-15. Parsed locally, 2026-08-17.

  2. 02

    Your row, found

    Matched on the name as filed. We recompute the rate from your own case counts and hours, and print both.

    Your company’s own row, public since the day it was filed.

  3. 03

    The two criteria, and your place

    Your rate against the national average, its direction across CY2021 to CY2023, and your percentile among the filers in your NAICS code.

    The same file, plus filing years back to CY2016.

04what it buys

$149 a month reruns both criteria.

Sample

  • ALERT · Case #114, logged today as restricted duty
  • Rolling DART moves 5.9 → 7.2
  • Manufacturing benchmarks: median 0.88 · p90 5.80 · p95 8.28 (n = 43,501)
  • Both computable criteria rerun at the new rate

The benchmarks are real; the case is invented. Log a case and the rate moves that day, with an Area Office inspection watch and a readiness binder from OSHA’s citation-frequency data.

05the pushback

Your DART rate is only half the readout.

I already know my DART rate.
You calculated it and signed it. The percentile and the trend test are OSHA’s work, and OSHA sends neither.
This is fear-mongering.
Every figure here prints its source and its row count, and the board names the two criteria nobody outside OSHA can compute.
My carrier already benchmarks this.
Of nine loss-control programs, four could be read directly (EMPLOYERS, ICW Group, AmTrust, Travelers), a fifth only through search (Texas Mutual), and none publishes a peer percentile. Four would not load; ask yours.
A free calculator already does this.
SafetyEvolution and BLS both publish one, free and ungated. They compare your rate to an industry average; this ranks your establishment among the filers in OSHA’s own file.

One logged case changes your rate and your percentile.

$149 a month if it ships

Every figure comes from OSHA’s published file or your own filing, and the two unpublished cutoffs stay marked unknown.