Recordkeeping: OSHA 300 Forms and Safety Metrics

The Three OSHA Recordkeeping Forms

OSHA recordkeeping lives in 29 CFR Part 1904 (note: 1904, not 1926 — recordkeeping is one of the few general-industry parts that applies to construction). Three forms work together:

The OSHA 200 form was the predecessor to the 300 and was replaced in 2002. A distractor naming the 200 form is testing whether you know it is obsolete.

Posting and Retention Rules

Two numbers are heavily tested:

• Posting: The 300A summary must be posted in a conspicuous workplace location from February 1 to April 30 of the year following the records — three months. Only the 300A is posted, never the full 300 Log (which contains private medical detail).
• Retention: Under 29 CFR 1904.33, you must keep the 300 Log, 300A summary, and 301 reports for 5 years following the calendar year they cover, and update the 300 Log if cases change.

Who Must Keep Records

Not every employer keeps OSHA logs:

• Employers with 10 or fewer employees during all of the prior calendar year are partially exempt from routine recordkeeping (1904.1).
• Employers in certain low-hazard industries (Appendix A) are exempt — but construction (NAICS 23) is NOT on that list, so construction employers above 10 employees must keep records.
• Electronic submission (Injury Tracking Application): construction establishments with 20-249 employees submit the 300A electronically; establishments with 100+ employees in designated high-hazard industries submit the 300, 300A, and 301.

Critical exception: Even a partially exempt small employer must still report fatalities and severe injuries to OSHA. Recordkeeping exemption does NOT waive the reporting duty.

Severe-Injury Reporting Deadlines (1904.39)

Memorize these two deadlines cold:

• Fatality — report within 8 hours of learning of it.
• In-patient hospitalization, amputation, or loss of an eye — report within 24 hours.

Reports go to OSHA by phone (1-800-321-OSHA), the nearest area office, or the online reporting portal. Mixing up 8 and 24 hours is a classic trap.

What Makes an Injury Recordable (1904.7)

A case is recordable if it is work-related and results in any of the following general recording criteria:

• Death
• Days away from work
• Restricted work or job transfer
• Medical treatment beyond first aid
• Loss of consciousness
• A significant injury/illness diagnosed by a physician (e.g., fractured bone, punctured eardrum, cancer)

The line between first aid (not recordable) and medical treatment beyond first aid (recordable) is decisive. OSHA gives a closed list of what counts as first aid — if a treatment is not on the list, it is medical treatment and the case is recordable. First aid includes:

• A single dose of non-prescription medicine at non-prescription strength.
• Cleaning, flushing, or soaking wound surfaces; using bandages, butterfly closures, or gauze (but not sutures).
• Hot or cold therapy, elastic bandages, and non-rigid supports.
• Drilling a fingernail to relieve pressure, or draining a blister.
• Drinking fluids for heat stress relief.

By contrast, sutures, prescription medication, rigid splints/casts, physical therapy, or any restricted duty push a case into recordable territory.

Work-Relatedness

Before recordability, the case must be work-related — caused or aggravated by an event or exposure in the work environment. OSHA presumes work-relatedness for injuries occurring on the job, subject to specific exceptions (e.g., the common cold/flu, eating one's own food, voluntary wellness activities). On the STSC, if a case is not work-related, it is not recorded regardless of severity.

The Two Core Safety Metrics

Lagging indicators measure outcomes. The STSC expects you to know both formulas, including the constant 200,000, which represents 100 full-time employees working 40 hours/week for 50 weeks.

• TRIR (Total Recordable Incident Rate): TRIR = (Number of recordable cases x 200,000) / Total hours worked
• DART (Days Away, Restricted, or Transferred): DART = (cases with days away + restricted/transferred x 200,000) / Total hours worked

Worked example: A company logs 5 recordable cases in a year with 250,000 hours worked. TRIR = (5 x 200,000) / 250,000 = 1,000,000 / 250,000 = 4.0. If 2 of those 5 cases involved days away or restriction, DART = (2 x 200,000) / 250,000 = 1.6.

Reading the Numbers in Context

A TRIR by itself means little until you compare it. The Bureau of Labor Statistics (BLS) publishes industry-average rates, and construction firms benchmark their TRIR and DART against that average and against their own prior years. A rate that is trending down signals an improving program; a sudden spike signals a control or training breakdown worth investigating. Many owners and general contractors set a TRIR ceiling as a prequalification requirement for bidding work, so the metric has commercial weight, not just regulatory weight.

Be careful with the limits of lagging metrics: a small company with few hours worked can show a volatile rate, because a single recordable case swings the number sharply. That is one more reason mature programs pair lagging rates with leading indicators that are less subject to small-sample noise.

Leading vs. Lagging Indicators

• Leading indicators are proactive: inspections completed, near-miss reports filed, training hours, safety observations, JHA completion. They predict and prevent.
• Lagging indicators are reactive: TRIR, DART, lost-time injuries, fatalities. They count harm that already happened.

A mature program tracks both but emphasizes leading indicators because they create the chance to intervene before an injury. A scenario asking which metric is proactive wants a leading indicator; one asking which measures outcomes wants a lagging indicator.

Common Recordkeeping Mistakes

• Confusing the 8-hour fatality deadline with the 24-hour hospitalization/amputation/eye-loss deadline.
• Posting the full 300 Log instead of just the 300A summary.
• Assuming a small or exempt employer need not report a fatality (they must).
• Forgetting the 200,000 constant or dividing by the wrong figure in TRIR/DART.
• Treating TRIR as a leading indicator — it is lagging.