Atmospheric Testing for Confined Spaces: Order, Technique and Continuous Monitoring

Most confined space victims never knew the air was bad. Testing replaces guesswork with numbers, but only if it's done in the right order, in the right places, with an instrument that's working.

When testing is required

Under 1910.146(d)(5), the employer must:

• Test conditions before entry is authorized to determine whether acceptable entry conditions exist.
• Test or monitor as necessary during entry to make sure acceptable conditions are maintained.
• Test in a specific order: oxygen first, then combustible gases, then toxic gases and vapors.
• Give authorized entrants or their representatives the opportunity to observe pre-entry and later testing, and retest when an entrant requests it.

Testing must use a calibrated direct-reading instrument.

Why oxygen, then flammables, then toxics

1. Oxygen first. Many combustible gas sensors need enough oxygen to read correctly. In an oxygen-deficient atmosphere, the flammable reading may be falsely low. Low oxygen is also immediately dangerous.
2. Flammables second. A fire or explosion threatens everyone at once, including people outside the space, and happens instantly.
3. Toxics last. Toxic gases are just as deadly, but most cause harm through exposure over time or at higher concentrations, so they're checked after the more immediate threats.

Typical 4-gas monitor readings

Most entries use a 4-gas monitor reading oxygen, flammable gas as % LEL, hydrogen sulfide and carbon monoxide. Common default alarm setpoints:

Your entry permit's acceptable entry conditions are what count. Many programs use the monitor's alarm setpoints as the limits. Check readings with our gas reading checker.

Testing a stratified space

Gases layer by weight. Methane rises toward the top, hydrogen sulfide and many solvent vapors settle low, and carbon monoxide mixes roughly evenly. Test the whole space, not just the opening:

• Test from outside before anyone enters, using a sampling pump and probe or hose.
• Test the top, middle and bottom of the space. Appendix B to 1910.146 recommends testing about 4 feet in the direction of travel and to each side as you move through a stratified space.
• Allow for sampling time. A pump draws air through the hose at a set rate, so add time for the sample to reach the sensor, generally a few seconds per foot of hose, per the manufacturer.
• Record results, times and the tester's name or initials on the permit.

Bump tests and calibration

A monitor that hasn't been checked can read zero in a deadly atmosphere. Two checks keep it honest:

• Bump test (functional test): briefly exposes the sensors to a known gas to confirm they respond and the alarms go off. It doesn't adjust the instrument.
• Calibration: adjusts the instrument's readings to match a known concentration of test gas.

OSHA recommends verifying calibration with a bump test or calibration check before each day's use, following the manufacturer's instructions, and calibrating whenever a monitor fails a bump test. Learn more in our bump test vs. calibration guide.

Continuous monitoring

Conditions change. Work inside the space, process leaks, ventilation failure and weather can all shift the readings. Good practice, and the construction standard, call for monitoring throughout entry:

• In construction, 1926.1204 requires the employer to continuously monitor atmospheric hazards unless it can show that continuous monitoring equipment isn't commercially available or that periodic monitoring is enough.
• In general industry, 1910.146 requires testing or monitoring as necessary to make sure conditions stay acceptable. Most programs have entrants wear a personal monitor and keep a monitor at the work area.
• Under alternate procedures, the atmosphere must be tested periodically, and any hazardous atmosphere means immediate exit.

Key takeaways

• Test before entry and as needed during entry with a calibrated direct-reading instrument.
• Order: oxygen, then flammable gases, then toxic gases and vapors.
• Test from outside at several levels, allowing time for the sample to reach the sensor.
• Bump test before each day's use and calibrate when needed.
• Monitor continuously during entry, required in construction unless not feasible or periodic monitoring is sufficient.

Last reviewed: October 25, 2026 · Free general education based on OSHA 29 CFR 1910.146 and 29 CFR 1926 Subpart AA. It does not replace the site-specific training, written program and hands-on evaluation your employer must provide, and it does not issue a certificate. Nothing you enter on this page is saved or sent.