☎ (888) 264-7050
Your Respiratory Protection Program Under 1910.134: What Has to Be in Writing, and the Part OSHA Proposed to Delete
← Back to Blog Industrial Hygiene

Your Respiratory Protection Program Under 1910.134: What Has to Be in Writing, and the Part OSHA Proposed to Delete

29 CFR 1910.134 requires nine written program elements. OSHA proposed deleting the most cited one, and its own advisory committee voted 5 to 4 against it.

by iSi Environmental

Your safety manager can produce the respirators. That was never the question. When an OSHA compliance officer asks for the written program, the medical clearance on file for the maintenance tech who wears a half mask twice a year, and the fit test record that matches the facepiece he is actually wearing, who produces those?

That gap is what 29 CFR 1910.134 is built to expose. The standard is organized around one idea most facilities have backwards: the respirator is not the control, the program is. Paragraph (a)(1) says it plainly, that engineering controls come first and respirators are what you use when engineering controls are not feasible or are being installed. Everything downstream, selection in (d), medical evaluation in (e), fit testing in (f), use in (g), maintenance in (h), breathing air in (i), cartridge identification in (j), training in (k), program evaluation in (l) and recordkeeping in (m), exists to make one written document defensible.

⚡ TL;DR: 1910.134(c)(1) requires a written, worksite-specific program covering nine enumerated elements. There is no annual review requirement in the rule. The standard sets only two recurring clocks, both per employee: an annual fit test and annual retraining. And the single element OSHA cites most is the one it has proposed to delete, with no final rule issued.

⚠️ What’s at stake: For penalties assessed after January 15, 2026, OSHA’s maximum is $16,550 per serious or other-than-serious violation and $165,514 per willful or repeated violation (OSHA Penalties). Program gaps do not cite as one item. No written program, no medical evaluation, no fit test and no training are four separate duties under (c)(1), (e)(1), (f)(2) and (k), so they are four separate exposures from one inspection.

The most contested rule in OSHA’s current docket is this one, and it is moving in two directions at once

Every other page on this topic treats 1910.134 as settled. It is not. On July 1, 2025 OSHA published a proposed rule at 90 FR 28463 (RIN 1218-AD48, Docket OSHA-2025-0006) that would add a new (e)(1)(ii) exempting employees required to use filtering facepiece respirators and loose-fitting powered air-purifying respirators from medical evaluation entirely. OSHA’s own economic analysis puts 2,733,069 private-sector workers in those respirator types across 131,089 firms, 127,351 of them small under SBA standards, and claims roughly $75 million per year in savings.

It is still a proposal. OSHA’s Advisory Committee on Construction Safety and Health voted 5 to 4 on April 1, 2026 to recommend OSHA not proceed. OSHA then reopened the record at 91 FR 33130 specifically to take comment on that recommendation, held a hearing in August 2026, closed post-hearing comments on September 19, 2026, and on October 28, 2026 was still being urged by 28 members of Congress to abandon the rulemaking. That letter cites, among others, the Associated General Contractors of America arguing that quality safety rules save contractors money over time. An employer trade association arguing against its own deregulation is worth noticing.

Now the direction nobody is writing about. OSHA’s deregulatory rulemaking docket holds 23 pending proposals, all at the proposed-rule stage with none final, and 17 of them touch respiratory protection. Sixteen are substance-specific standards, including asbestos, benzene, cadmium, coke oven emissions, ethylene oxide, formaldehyde, inorganic arsenic, lead, methylene chloride and vinyl chloride, whose own respirator provisions would be removed as duplicative of 1910.134. If those finalize, your 1910.134 program becomes the single governing respirator rule for every one of those substances. The written document gets more load-bearing, not less. The related 13-carcinogens and respiratory proposal is covered separately in OSHA’s 2026 proposed rule changes on carcinogens and respiratory protection.

The practitioner consequence runs opposite to the headline. Nothing in the current standard has changed. Every obligation in (e) is enforceable today. A facility that already stopped medical evaluations for its N95 users in anticipation of a rule that has not issued is out of compliance right now.

What has to be in an OSHA written respiratory protection program?

Nine elements, enumerated at 29 CFR 1910.134(c)(1)(i) through (ix): selection procedures, medical evaluations, fit testing, proper use in routine and emergency situations, cleaning and storage and maintenance schedules, breathing air quality for supplied-air units, hazard training, use training including donning and doffing, and procedures for regularly evaluating program effectiveness. All must be worksite-specific and in writing.

Worksite-specific is the word that does the work. A binder bought off a template and never adapted fails on its face, because the procedures in it do not describe your contaminants, your respirator models or your facility. An element may be left out only if it genuinely does not apply at that location, and the absence of supplied-air respirators is the only common case. Under the current text at eCFR, the program is the affirmative deliverable. Without it, OSHA’s working assumption is that you have none.

Who can serve as a respiratory protection program administrator?

1910.134(c)(3) requires someone “qualified by appropriate training or experience that is commensurate with the complexity of the program.” OSHA mandates no certification and no specific course. Basic knowledge can suffice for nuisance dust. IDLH atmospheres, supplied air, or multiple contaminants demand more. Name the person or position in the written program and keep the evidence of their qualification.

This is a performance requirement, not a credential requirement, and that cuts both ways. Nobody can tell you your administrator is unqualified because they lack a certificate. Equally, nobody will accept an administrator whose only qualification is that the job landed on them. The practical test is whether that person can explain, without reading from the binder, why each respirator in the building was selected and what would change the selection.

Do you need a written respiratory protection program if respirator use is voluntary?

Not if the only respirators are filtering facepieces. Under 1910.134(c)(2), where use is not required and the employer has determined voluntary use creates no hazard, dust-mask users can be left out of the written program. The surviving obligation is Appendix D information, required by (k)(6). Voluntary use of elastomeric or tight-fitting PAPRs does trigger the written program.

This is the exemption most often claimed and most often blown. The moment a voluntary user reaches for an elastomeric half mask instead of a dust mask, the full program obligation attaches to that person, including medical evaluation and cleaning and storage procedures. The determination that voluntary use creates no hazard is also an affirmative finding you should be able to show, not a default. The medical evaluation side of voluntary use is worked through in detail in OSHA’s medical evaluation requirements for voluntary respirator use.

How often does a respiratory protection program have to be reviewed?

There is no interval in the rule. 1910.134(l)(1) requires workplace evaluations “as necessary” to confirm effective implementation, and (l)(2) requires regular consultation with respirator wearers about fit, selection, real-world use and maintenance. Annual review is a common practice, not a legal requirement. Workplace or respirator changes are the real trigger.

Facilities get this wrong in both directions. Some spend budget every January rewriting a document that nothing has changed. Others never touch it after a process change, a new contaminant or a switch of respirator model, each of which is an actual trigger. The (l)(2) consultation duty is the part that gets skipped entirely: asking the people who wear the respirators about breathing difficulty, restricted motion, interference with vision or communication, and discomfort, then correcting what those conversations surface. That record is also the cheapest evidence you have that the program is more than paper.

Does the medical evaluation have to come before fit testing?

Yes. 1910.134(e)(1) requires the medical evaluation before the employee is fit tested and before first required workplace use. The sequence is the compliance point: a fit test performed on a medically uncleared employee is a defective record. OSHA has a pending proposal to narrow this for filtering facepieces, but no final rule exists as of November 2, 2026.

Order matters because the two steps answer different questions. The evaluation asks whether this person can safely wear a respirator at all; the fit test asks whether this facepiece seals on this face. Run them backwards and the fit test record proves nothing, because the premise it rests on was never established. Seasonal and temporary workers are inside this requirement if their jobs require respirator use. The mechanics of the evaluation itself, the Appendix C questionnaire, the PLHCP review, and what the employer may and may not see, are covered in OSHA respiratory protection medical evaluation requirements.

Who pays for respirators, fit testing, and medical evaluations?

The employer. 1910.134(c)(4) requires respirators, training and medical evaluations to be provided at no cost to the employee. The one exception sits in voluntary use: employers need not pay for filtering facepieces employees choose to wear, but must still cover medical evaluations for voluntary users of other respirator types and provide time and facilities for maintenance.

The provision is short and gets argued about anyway, usually over cost-sharing on elastomeric masks assigned to individuals or over whether fit testing on an employee’s own time is acceptable. It is not. Time spent on required medical evaluation, fit testing and training is employer time, and the equipment is employer-provided.

How long do respirator fit test and medical records have to be kept?

Fit test records must be retained only until the next fit test is administered, per 1910.134(m)(2)(ii), which in an annual program means roughly one year. Medical evaluation records follow 29 CFR 1910.1020 instead, a far longer obligation. A current written copy of the program must be retained under (m)(3).

The asymmetry surprises people. The document that feels most like the compliance record, the fit test, has the shortest life, while the medical file sits under the general exposure and medical records rule with its much longer retention. In practice, keeping fit test records longer than the minimum is worth doing anyway, because a multi-year history is what demonstrates a program rather than a single event. The five data fields required at (m)(2)(i) are the part to get right: who, which respirator, which protocol, the date, and the results.

What does OSHA require if you cannot estimate the airborne exposure?

1910.134(d)(1)(iii) is unforgiving here: where the employer cannot identify or reasonably estimate employee exposure, the atmosphere must be considered IDLH. That default drives selection up to a pressure-demand SCBA or a supplied-air respirator with escape provisions. The cheaper path is almost always to sample and document the exposure rather than accept the IDLH presumption.

This is the paragraph with the largest dollar consequence in the standard, and almost nobody reads it as a cost provision. Treating an atmosphere as immediately dangerous to life or health is not a paperwork posture. It dictates the respirator class, the standby and rescue provisions under (g)(3), and the training that goes with them. An exposure assessment that puts a number on the contaminant usually costs a fraction of standing up a supplied-air program you did not need. Breathing resistance and selection tradeoffs across respirator classes are worked through in respirator pressure drop and selection.

When is respirator retraining required beyond the annual cycle?

1910.134(k)(5) requires retraining annually and on three triggers: workplace or respirator-type changes that render prior training obsolete, observed inadequacies in an employee’s knowledge or use, and any other situation where retraining appears necessary for safe use. A new hire trained elsewhere within the last 12 months can be credited under (k)(4) if knowledge is demonstrated.

Training is paragraph (k), not (j). Paragraph (j) is the identification of filters, cartridges and canisters, and the two get swapped in citations and in program documents often enough to be worth saying out loud. The content requirements sit at (k)(1)(i) through (vii) and training must come before first workplace use under (k)(3). The observed-inadequacy trigger is the one with teeth: if a supervisor watches someone don a respirator wrong and nothing happens afterward, the retraining duty was triggered and missed.

What is the penalty for a respiratory protection program violation?

For penalties assessed after January 15, 2026, OSHA’s maximums are $16,550 per serious, other-than-serious or posting violation, $16,550 per day beyond the abatement date for failure to abate, and $165,514 per willful or repeated violation. Program gaps tend to cite as separate items: no written program, no medical evaluation, no fit test and no training are four exposures, not one.

Those figures are numerically identical to the January 2025 maximums, and that is itself the story. The 2026 inflation adjustment was cancelled rather than applied, which we covered in the 2026 federal penalty freeze. The flat schedule is not relief. It means the same multiplied exposure as last year with no offsetting change, and the multiplication is what matters here. One inspection at a facility with respirators and no program does not produce one citation.

Where does respiratory protection rank among OSHA’s most-cited standards now?

Sixth. OSHA’s preliminary FY 2026 list, covering October 1, 2025 through August 31, 2026, ranks 1910.134 behind fall protection (1926.501), hazard communication (1910.1200), lockout/tagout (1910.147), scaffolding (1926.451) and ladders (1926.1053). That is a change from the FY 2021 picture, when respiratory protection was the top general-industry citation.

Worth correcting the record, because the earlier version of this post reported the FY 2021 standings and the order has since reversed: hazard communication is now second, not fifth. Note also what OSHA does not publish. The top 10 list gives rankings without citation counts, and no primary OSHA source breaks 1910.134 down by subparagraph. Anyone telling you which element of the standard is cited most often is not working from published data.

Do Kansas, Oklahoma, Texas, Missouri or Nebraska have their own respirator rules?

No. None of the five has an OSHA-approved State Plan, so federal 1910.134 applies directly to private-sector employers in all of them with no state overlay and no more-stringent state respirator standard to reconcile. Public-sector employees in those states fall outside federal OSHA coverage, which is a gap, not a separate standard.

OSHA’s State Plans page lists 29 approved plans, none covering Kansas, Oklahoma, Texas, Missouri, Nebraska or Colorado. The practical consequence is worth naming: a municipal water plant or a county highway department in Kansas operates under no enforceable federal respirator standard. Many adopt 1910.134 voluntarily or because an insurer requires it, which means the program document exists but has never been tested by anyone.

What Region VII’s health hazard emphasis program actually asked for

What varies across iSi’s service region is enforcement posture, not rule text. OSHA Region VII, covering Kansas, Missouri, Iowa and Nebraska, has run CPL 2-21-001, a Top 50 Workplace Health Hazard emphasis program that explicitly tracks 1910.134 alongside 1910.95, 1910.120, 1910.132, the 1910.1000 series and the 1910.1025 through 1450 substance standards. It directs compliance officers to pull exposure monitoring records and OSHA 300 logs at the opening conference, to screen for full-shift sampling opportunities during the walkaround, and to sample as soon as practicable. Inspections under it are coded T50WHH, and the 50-NAICS appendix was ranked by serious health violations per inspection, which is why it reaches industries most people would not predict, including lessors of residential buildings, pottery and plumbing fixture manufacturing, dental offices and nail salons.

One fact and one uncertainty, stated as both. The directive’s own face carries an effective date of October 1, 2021 and an expiration date of September 30, 2026, roughly five weeks before this update. OSHA still lists it on its regional directives page with no cancellation note, and we found no successor directive. What that means for targeting in 2027 is not something anyone can state from published sources yet. What it does tell you is the shape of a health-hazard inspection in this region: exposure records first, sampling second.

What to do with this

Four checks, in the order that finds problems fastest.

  1. Open the written program and confirm all nine elements at (c)(1)(i) through (ix) are present and describe your facility, not a template’s.
  2. Pick one respirator wearer and trace them end to end: medical clearance dated before the fit test, fit test matching the facepiece they actually wear, training within 12 months, and a signature.
  3. List every voluntary user and confirm which respirator type each wears. Anyone in anything other than a filtering facepiece belongs inside the written program.
  4. Write down what triggered your last program review. If the answer is the calendar rather than a workplace change, the (l) duty is being performed backwards.

If check 2 or check 3 turns up a gap, the exposure assessment is usually the piece to get right first, because (d)(1)(iii) decides the respirator class before anything else does. iSi writes and reviews written respiratory protection programs, runs the exposure monitoring that keeps a facility out of the IDLH presumption, handles fit testing and respirator selection, and conducts the (k) training.

Need your respiratory protection program reviewed against the current standard? Contact iSi Environmental for a same-day response, or explore our safety consulting services. We Plug In. You Level Up.

Frequently Asked Questions

What has to be in a written respiratory protection program?

Nine elements enumerated at 29 CFR 1910.134(c)(1)(i) through (ix), covering selection, medical evaluation, fit testing, use, maintenance, breathing air quality, training and program evaluation. The program must be worksite-specific and administered by a qualified program administrator.

Is an annual respiratory protection program review required?

No. 1910.134(l) requires workplace evaluations as necessary plus regular consultation with respirator wearers, with no stated interval. The two fixed annual clocks in the standard are per-employee: fit testing and retraining.

Do voluntary dust-mask users need to be in the written program?

Not if filtering facepieces are the only respirators involved and the employer has determined voluntary use creates no hazard; those users still must receive Appendix D information. Voluntary use of elastomeric respirators or PAPRs does trigger written program elements.

Did OSHA eliminate medical evaluations for N95 users?

No. The July 2025 proposal at 90 FR 28463 would exempt required users of filtering facepieces and loose-fitting PAPRs from medical evaluation, but it is not final, the record was reopened in June 2026, and every obligation in 1910.134(e) remains enforceable today.

Sources

Regulatory citations relied on: 29 CFR 1910.134(a) through (n); 29 CFR 1910.1020; 29 CFR 1903.15; 29 CFR 1915.154; 29 CFR 1917.92; 29 CFR 1918.102; 29 CFR 1926.103; 90 FR 28463; 91 FR 33130.

Tags: