Leading and Lagging Safety Indicators: OSHA Only Uses One of Them to Pick Your Inspection
← Back to Blog Safety

Leading and Lagging Safety Indicators: OSHA Only Uses One of Them to Pick Your Inspection

Leading indicators live in OSHA guidance. OSHA's Site-Specific Targeting program picks inspections from the lagging 300A injury data you submitted yourself.

by iSi Environmental

OSHA has already scored your facility. It used a number you calculated, certified and submitted yourself, and it never asked about a single leading indicator. Safety conferences treat proactive and reactive metrics as a philosophy argument, where leading indicators track the work that prevents injuries and lagging indicators count the injuries that already happened. OSHA’s main programmed inspection program for general industry does not participate in that argument. It selects establishments from the injury and illness data employers submit under 29 CFR 1904.41, and from nothing else.

There is a date attached to this. OSHA’s April 16, 2024 enforcement memo fixes the limitations period for citing a failure to submit at six months from the March 2 due date, and says so plainly: “The six-month date to issue a citation for non-compliance with the requirements of 29 CFR 1904.41 will therefore be September 2” (OSHA memo, 4/16/2024). For calendar year 2025 data due March 2, 2026, that window closes September 2, 2026. An establishment that missed the submission is inside a live enforcement window right now.

What is the difference between leading and lagging safety indicators?

Lagging indicators count events that already occurred: injuries, illnesses, fatalities, DART and TRIR rates. Leading indicators are proactive, preventive and predictive measures of safety activity, such as training completion, inspections closed, or maintenance performed on schedule. OSHA draws this distinction in publication OSHA 3970 (June 2019). Neither category is required by any OSHA standard.

That last sentence is the one that gets skipped. OSHA 3970 is guidance. It carries no citation authority and creates no record an inspector can demand. An employer with a spotless leading-indicator dashboard and an employer with no metrics program at all are in identical regulatory positions under Part 1904.

Does OSHA use leading indicators to decide who gets inspected?

No. OSHA’s Site-Specific Targeting program runs entirely on lagging data. CPL 02-01-067 selects establishments using Form 300A injury and illness data submitted under 29 CFR 1904.41 for CY2021 through CY2023. No leading indicator appears anywhere in the selection criteria. Leading indicators matter after the inspector arrives, not before.

The directive states its own basis in the Background section: the SST plan “uses objective data from injury and illness information that employers submit under 29 CFR § 1904.41.” Signed April 8, 2025 and effective May 20, 2025, CPL 02-01-067 terminates May 20, 2027 unless a new instruction replaces it. It is operative today.

Leading-indicator work does have a place. Section XII.E of the directive tells the compliance officer to evaluate whether the establishment’s safety and health management system is adequate to address the elevated rate that put it on the list. Documented training, inspection and correction activity becomes defensible evidence at that point. It changes the conversation once the inspector is on site, and does nothing about whether the inspector shows up.

How does OSHA use my Form 300A data to select my facility for inspection?

OSHA builds four lists from submitted 300A data: establishments with elevated CY2023 DART rates, establishments trending upward across CY2021 to CY2023, establishments that failed to submit CY2023 data, and a random sample of low-DART establishments used to verify data accuracy. Area Offices then draw inspection cycles of 5 to 50 establishments from those lists.

Cycles above 50 establishments require Regional Office approval (§X.B), and the program reaches non-construction general industry establishments with 20 or more employees. Once an inspection opens, the compliance officer reviews the 300 logs, 300A summaries and 301 reports for CY2021 through CY2023 under §XII.C. Three years of records have to reconcile with each other and with the numbers already sitting in OSHA’s database.

What DART rate puts my establishment on OSHA’s SST list?

OSHA does not publish a single number. CPL 02-01-067 §IX.A.1 sets one DART threshold for manufacturing (NAICS 31-33) and a different one for all other non-construction NAICS, so the two groups are targeted proportionally. For the upward-trending list the directive is explicit: at or above twice the private sector national average in CY2022, still rising through CY2023.

A manufacturer is judged against a manufacturing-specific threshold, so a rate that looks reasonable next to all-industry data may still be elevated inside its own NAICS group.

Can a low injury rate get my facility inspected?

Yes, and most employers do not know it. CPL 02-01-067 §IX.A.3 directs OSHA to generate a random sample of establishments with low DART rates using CY2023 data, for the stated purpose of verifying the reliability of reported 300A data. A clean number invites a quality-control check, so the 300 log has to substantiate the summary.

This inverts how most safety committees read their own metrics. A DART rate well below the industry average gets presented to ownership as proof the program is working. OSHA reads the same number as a data-integrity question worth sampling.

The exposure is not the low rate itself. It is a low rate sitting on top of recordable cases classified as first aid, restricted-duty days that were never counted, or a log reconstructed at year end from memory. Area Directors may order a full recordkeeping audit where there is evidence of systemic problems.

What happens if I do not submit my 300A data to OSHA?

Non-submission puts the establishment on the SST non-responder list, from which OSHA draws a random sample for programmed inspection. The directive states the purpose plainly: to discourage employers from withholding data to avoid inspection. Failure to submit is citable under 29 CFR 1904.41, carrying up to $16,550 per violation as an other-than-serious citation (OSHA memo, 5/21/2026).

Willful or repeat violations run to $165,514 per violation. Neither figure moved this year. OSHA carried the 2025 penalty amounts forward into 2026 because no October 2025 CPI-U was published, so the usual January inflation adjustment did not occur.

There is a defense built into the program, and it works on the spot. Under §X.D.3, if a compliance officer arrives citing non-responder status and the establishment produces documentation that the submission was in fact made, the inspection is terminated and coded “No Inspection.” Same outcome if the establishment can show it is not subject to the requirement by size or industry. Keep that confirmation where the front office can find it in five minutes.

How long does OSHA have to cite me for failing to submit 300A data?

Six months from the March 2 deadline, which OSHA’s April 16, 2024 enforcement memo fixes at September 2 of the same year. For CY2025 data due March 2, 2026, the citation window closes September 2, 2026. Compliance officers are directed to check the ITA database during all inspections, not only SST-targeted ones.

The memo’s language on that last point matters more than the deadline does. Area Offices are told that compliance officers “should refer to this database during all inspections to identify employers that were required to submit records but failed to do so.” Any inspection, from any source, opens with a database check.

Which establishments have to submit injury data to OSHA electronically?

Three groups under 29 CFR 1904.41: establishments with 250 or more employees subject to the recordkeeping rule; establishments with 20 to 249 employees in an Appendix A (Subpart E) industry; and establishments with 100 or more employees in an Appendix B (Subpart E) industry, which submit Forms 300 and 301 in addition to 300A. Under 20 employees at all times means no routine submission.

The 20-employee floor is real, and it is worth stating because it gets misreported. OSHA’s memo puts it directly: “Establishments with fewer than 20 employees at all times during the year do not have to routinely submit information electronically to OSHA.” An Appendix A industry classification does not by itself create a filing obligation for a 12-person shop.

Headcount is where facilities miscount. Part-time, seasonal and temporary workers all count, and each individual employed at any point during the calendar year counts as one employee. A plant that runs 15 people year round and adds a summer crew can clear 20 without anyone noticing. Thresholds apply per establishment, since 1904.30 requires a separate 300 log for each site expected to operate a year or longer.

The submission mechanics themselves are covered in our post on OSHA Form 300A submission deadlines and electronic recordkeeping. Everything downstream of the March 2 filing is what this article is about.

Does OSHA’s Site-Specific Targeting program apply to construction?

No. CPL 02-01-067 §I states the SST program is OSHA’s main site-specific programmed inspection initiative for general industry workplaces and does not include construction worksites. Construction is targeted through separate mechanisms. Public sector employers are also deleted from SST lists, with the U.S. Postal Service the standing exception.

For a manufacturer with a construction subsidiary, the plant sits in the SST population and the construction arm does not.

Do Kansas, Oklahoma, Missouri, Nebraska and Texas follow the federal SST program?

All five are under federal OSHA jurisdiction for private sector employers, with no OSHA-approved State Plan, so CPL 02-01-067 applies directly and without state variation. State Plan states must adopt targeting policies at least as effective as the federal SST and notify OSHA within 60 days, but that mechanism does not affect iSi’s core region.

State and local government workers in these five states are not covered by federal OSHA.

How do I get removed or deferred from OSHA’s SST inspection list?

Four routes exist in the directive. A comprehensive safety or health inspection within the previous 36 months triggers deletion. VPP participants are deleted for the duration of participation. SHARP participants are deferred up to 2 years initially and up to 3 on renewal. A scheduled initial comprehensive On-Site Consultation visit defers a programmed inspection up to 90 calendar days.

The On-Site Consultation route to SHARP is the one most manufacturers in this region have never priced out. It is a free state-run program, separate from enforcement, and SHARP status defers programmed inspections for up to two years on initial approval under §X.E.1 and 29 CFR 1908.7(b)(4)(i). Pre-SHARP defers up to 18 months. For a facility that knows its rate is elevated and knows why, that is a legitimate path off the list.

Which leading indicators actually hold up under research scrutiny?

Fewer than the marketing suggests. A 2023 Journal of Safety Research synthesis of the leading-indicator literature found the term’s definition, application and function “mostly ambiguous and inconsistent,” and concluded that lagging indicators remain the most appropriate evidence base from which leading indicators are built. Indicators that are easy to score are easy to game.

The full citation is Bayramova, Edwards, Roberts and Rillie, “Constructs of leading indicators: A synthesis of safety literature,” Journal of Safety Research Vol. 85 (2023), pp. 469 to 484. It is worth reading before buying a metrics platform, because the peer-reviewed position is that lagging data is the foundation leading indicators get derived from, rather than the obsolete thing they replace.

A second line of criticism comes from trade opinion and review literature rather than primary regulation, and it is still worth knowing. Many common leading indicators measure activity that has already occurred: a near miss was reported, training was delivered, an inspection was completed. Rare events remain hard to predict at the establishment level, and organizations with mature dashboards still have catastrophic incidents. Much of the loudest advocacy comes from vendors selling the software that tracks these metrics.

None of this argues against measuring safety activity. It argues for knowing which of your numbers is load bearing. The one OSHA acts on is the one you already filed.

The Honest Odds, and Why the Lists Still Matter

Federal OSHA and its state partners have roughly 1,850 compliance officers for 130 million workers, about one officer per 70,000 workers, and ran 34,696 federal inspections in FY2024 against more than 8 million worksites (OSHA Commonly Used Statistics). The odds of a random visit are small, and any article that tells you otherwise is selling something.

Trade and legal press reported a decline of roughly 20% in inspections over a six-month stretch of 2025 and a sharp drop in inspector headcount, which prompted a February 2026 letter to the Department of Labor from Senator Warren and five other senators. Attribute that where it belongs: the six-month figure is reporting from the trade press, not an OSHA figure. OSHA’s own enforcement summary has since posted the full-year FY2025 total: 30,273 federal inspections, down 12.6% from FY2024’s 34,625.

Thin enforcement resources are the argument for paying attention to targeting. An agency with one officer per 70,000 workers does not inspect randomly. It spends what it has on lists built from data employers hand over voluntarily. Between April 7, 2023 and December 12, 2024, SST produced 652 inspections, and OSHA’s evaluation in §VIII reports that SST sites showed a higher rate of violations per inspection and a higher not-in-compliance rate than other non-construction programmed inspections. The serious-violation rate was comparable, so the targeting finds more violations rather than more dangerous ones. OSHA concluded the program “continues to be an effective means of using its enforcement resources.”

Our earlier post on OSHA’s inspection weighting system covers what an inspection is worth to OSHA once it opens. This one covers how you get picked in the first place.

What to Do Before September 2

Five things, in order of exposure.

  1. Confirm the CY2025 submission was actually transmitted through the Injury Tracking Application and print the confirmation. If it was not, the citation window closes September 2, 2026, and the establishment lands on the non-responder list regardless.
  2. Verify the DART rate you submitted for CY2023 and reconcile it against the 300 log line by line. That is the year driving the current elevated-rate and low-rate lists.
  3. Reconcile the 300 logs, 300A summaries and 301 reports for CY2021 through CY2023. Those are the three years a compliance officer reviews under §XII.C.
  4. Recount employees per establishment using every part-time, seasonal and temporary worker on payroll at any point in the year, then check the NAICS against Appendix A and Appendix B to Subpart E.
  5. Assemble the safety and health management system documentation an inspector would ask for under §XII.E. This is where leading-indicator records earn their keep.

If your team handles inspection readiness in house, our Wichita and Kansas City pages walk through OSHA inspection preparation in more detail.

Where iSi Fits

Most facilities that end up on an SST list had a recordkeeping and reconciliation problem nobody owned, because the safety manager was carrying environmental reporting, permit tracking and program documentation on top of the work that needs their judgment on site.

iSi’s OSHA compliance audit work reconciles the 300 logs, 300A summaries and 301 reports against what was submitted, tests whether the management system documentation would hold up under a §XII.E review, and identifies whether an On-Site Consultation route to SHARP is worth pursuing. If your CY2025 submission is uncertain or your CY2023 DART rate has never been reconciled against the log, start there.

Sources

Tags: