Failed Emissions Test

Read the report before doing anything else. An emissions failure is not one thing — it is at least five different tests that share a piece of paper, and each fails for different reasons with different fixes. Most modern vehicles get an OBD-II test: the analyzer plugs into the diagnostic connector and asks the computer three questions. Is the malfunction indicator lamp commanded on? Are there stored emissions-related fault codes? Have the readiness monitors finished running? A commanded MIL is an automatic fail regardless of how clean the exhaust actually is. So is a burned-out MIL bulb, because the analyzer also checks that the lamp illuminates at key-on. Readiness monitors are the most common surprise. These are the computer's self-tests for the catalyst, oxygen sensors, evaporative system, EGR and others. Clearing codes or disconnecting the battery resets them all to incomplete, and they only reset to complete after the vehicle has been driven through specific conditions of speed, load, temperature and fuel level. Clearing a code the night before a test is one of the most reliable ways to fail it. Older or exempted vehicles may get a tailpipe or dynamometer test measuring hydrocarbons, carbon monoxide and oxides of nitrogen directly. Many programs also run a gas cap or evaporative pressure test, and a visual inspection for missing or tampered emissions equipment. Which tests apply, which model years are exempt, how many incomplete monitors are tolerated, and what waivers exist all vary by state and sometimes by county. Your report tells you which one you failed. Fix that.

Safety

A failed test is a compliance problem, not usually a safety one — with exceptions. A rich-running engine dumping unburned fuel into the catalytic converter can overheat it enough to glow and become an ignition source under the car. High tailpipe carbon monoxide combined with any exhaust leak ahead of the passenger compartment can put CO in the cabin, which is genuinely dangerous and causes headache, drowsiness and confusion before you recognize it. If the car failed for high CO and you smell exhaust inside, drive with windows open and get the exhaust inspected before anything else. Also treat a strong raw fuel smell as a leak, not an emissions issue.

Narrow it down

Each question splits the possibilities. Answer them in order — by the end you should have a much shorter list than you started with.

Does the report say you failed for incomplete readiness monitors rather than for a code or a gas reading?
If yes
Nothing may be wrong with the car. The computer's self-tests have not finished, almost always because the battery was disconnected or codes were cleared recently. Drive the vehicle normally through varied conditions and retest — no repair required, and no parts should be sold to you for this.
If no
You failed on a real result — a code, a lamp, a gas measurement, or a visual check. That points at an actual fault to diagnose.
Was the check engine light on before the test?
If yes
That alone failed the OBD-II portion. The stored code names the system that must be repaired. Repair the underlying fault, then let the monitors complete before retesting — clearing the code and driving straight back will fail you for incomplete monitors instead.
If no
If the lamp was off and you still failed OBD-II, check whether the lamp bulb illuminates at key-on. A dead MIL bulb or a bulb removed to hide a fault is a fail in most programs.
Was the failure a catalyst efficiency code, P0420 or P0430?
If yes
Do not go straight to a converter. Converters are usually killed by something else — a misfire dumping raw fuel, oil or coolant consumption poisoning the substrate, or a long-running rich condition. Also verify the downstream oxygen sensor is actually healthy, since a lazy sensor produces the same code. Replacing the converter without finding what killed it usually buys you a second dead converter.
If no
Rule out catalyst efficiency as the target and diagnose whatever the report actually names.
If it was a tailpipe test, which reading was out of range — HC, CO, or NOx?
If yes
High HC means unburned fuel: misfire, weak ignition, low compression, or a failing catalyst. High CO means the mixture is too rich: fuel control, injectors, fuel pressure, or an air metering fault. High NOx means combustion is running too hot or too lean: restricted EGR flow, a cooling system running hotter than spec, excessive carbon buildup, or a lean condition. Each points to a different subsystem.
If no
If your program is OBD-only, gas readings do not appear on the report and the codes and monitors are the whole story.
Was the engine fully warmed up when it was tested?
If yes
The test conditions were valid and the result reflects the vehicle's actual state.
If no
A cold or barely warm engine runs enriched and reads high on HC and CO, and some monitors will not run at all. Drive at least fifteen to twenty minutes, including some highway speed, immediately before a retest.
Has the vehicle had an aftermarket exhaust, intake, tune, or deleted emissions component installed?
If yes
Visual inspection and OBD-II readiness both catch this. A missing catalytic converter, disconnected air injection, removed EGR, or a tune that disables monitors will fail regardless of how the engine runs. Restoring original equipment is the only path through the test.
If no
Focus on the mechanical or electrical fault named in the report rather than on modifications.

Likely causes and how to check each one

1
Readiness monitors incomplete after a recent battery disconnect or code clear
Very common
Every emissions self-test resets to incomplete when the computer loses power or has its memory cleared. They only return to complete after the vehicle has run a specific drive cycle — combinations of cold start, steady cruise, deceleration, idle time, and in the case of the evaporative monitor, a fuel level within a particular range and a cold soak. The evaporative and catalyst monitors are typically the slowest to set. How many incomplete monitors a program tolerates varies by state and by model year.
How to check
A basic scan tool reads I/M readiness status directly and lists each monitor as complete or incomplete. Drive the vehicle normally for several days including highway and city driving, keep the fuel tank between roughly a quarter and three quarters full to allow the evaporative monitor to run, and recheck. If a specific monitor never sets after a week of varied driving, there is likely an unresolved fault in that system.
2
Malfunction indicator lamp commanded on, or a stored emissions code
Very common
In an OBD-II test the lamp state and stored codes are the test. Any commanded MIL fails the vehicle outright, regardless of tailpipe cleanliness. A missing or burned-out MIL bulb also fails, because the analyzer verifies the lamp works at key-on. Pending codes do not usually fail a test on their own but tell you what is about to.
How to check
Scan for stored and pending codes and note the freeze frame data, which captures the conditions when the fault set. Confirm the check engine lamp illuminates during key-on bulb check and extinguishes after start. Repair the named fault, then verify the monitors complete before retesting.
3
Catalyst efficiency below threshold
Common
The computer compares upstream and downstream oxygen sensor activity to judge whether the converter is still storing and releasing oxygen. When downstream activity starts mirroring upstream, it sets a catalyst efficiency code. The converter is the part that fails, but rarely the part that caused it — misfires, oil or coolant consumption, a rich condition, or a leaking exhaust upstream of the sensor all destroy or fool converters.
How to check
Compare upstream and downstream sensor waveforms on a scan tool: the downstream signal should be relatively slow and steady while the upstream cycles rapidly. Check for misfire counters, fuel trims well away from zero, oil consumption, and coolant loss. Inspect the exhaust for leaks ahead of the downstream sensor. Only after those are clean does converter replacement make sense.
4
Rich or lean fuel mixture
Common
A mixture problem raises hydrocarbons and carbon monoxide on a tailpipe test and often sets lean or rich codes on an OBD test. Causes include a contaminated or drifting mass airflow sensor, vacuum leaks, incorrect fuel pressure, leaking or clogged injectors, an exhaust leak drawing air past an oxygen sensor, and failing oxygen or air-fuel ratio sensors. The engine may drive normally throughout.
How to check
Read short and long term fuel trims at idle and at steady cruise. Positive trims indicate the computer is adding fuel to correct a lean condition; negative indicates it is removing fuel from a rich one. Trims that worsen at idle and improve under load suggest a vacuum leak. Smoke-test the intake and check fuel pressure against the manufacturer's specification, which varies by vehicle.
5
Misfire
Common
A misfiring cylinder sends raw fuel straight into the exhaust. That spikes hydrocarbons on a tailpipe test, sets misfire codes on an OBD test, and progressively destroys the catalytic converter. Worn spark plugs, failing coils, injector faults, vacuum leaks, and low compression all produce it, and a mild misfire can be entirely unnoticeable from the driver's seat.
How to check
Read per-cylinder misfire counters on a scan tool — they accumulate even below the code threshold. If one cylinder stands out, test that cylinder's spark, injector, and compression. If misfires are spread across all cylinders, look for a fuel supply, mixture, or timing problem rather than a single component.
6
Evaporative emissions leak or gas cap fault
Common
The evaporative system captures fuel vapor from the tank and routes it to the engine to be burned. The computer periodically pressure- or vacuum-tests it. A loose, wrong, or worn gas cap, a cracked hose, a leaking purge or vent valve, or a rusted filler neck all set leak codes and can fail a pressure test. This affects emissions but usually not driveability or fuel economy in any way you would notice.
How to check
Confirm the gas cap is the correct part, clicks when tightened, and has an intact seal. Beyond that, a shop smoke-tests the evaporative system to find the leak. Note that a large leak and a small leak set different codes and point to different-sized failures. This monitor is also the slowest to complete after a code clear.
7
High oxides of nitrogen
Occasional
NOx forms when combustion runs hot. It is measured on tailpipe and dynamometer tests but is not directly monitored on most OBD-II gasoline systems, so it can fail a tailpipe test with no light on. Common contributors are restricted EGR flow or a stuck EGR valve, a cooling system running hotter than designed, heavy carbon buildup raising compression, a lean condition, and on diesels a fault in the selective catalytic reduction or diesel exhaust fluid system.
How to check
Check for EGR-related codes and command EGR flow with a scan tool while watching manifold pressure or short term fuel trim respond. Inspect EGR passages for carbon blockage — they clog far more often than the valve fails. Verify the cooling system reaches but does not exceed the manufacturer's operating temperature. On diesels, check DEF quality, level, and any aftertreatment codes.
8
Visual inspection or tampering failure
Occasional
Inspectors look for the emissions equipment the vehicle was built with. A missing or aftermarket catalytic converter that does not meet the applicable standard, a removed air injection system, disconnected EGR, deleted diesel aftertreatment, or a tune that suppresses monitors will all fail on sight or on data, no matter how the engine runs. Some programs also flag a diagnostic connector that has been damaged or made inaccessible.
How to check
Compare installed components under the vehicle against the emissions control information label under the hood, which lists the systems that vehicle was certified with. Ask whether any prior owner or shop installed a tune, and have the calibration verified against the manufacturer's current part number — many programs check this electronically.
9
Engine tested cold or under invalid conditions
Occasional
Emissions equipment only works at temperature. A catalytic converter that has not lit off does little, and an engine still in warm-up enrichment runs deliberately rich. Some monitors will not run at all until the engine reaches operating temperature. A vehicle driven a short distance to the station in cold weather can fail a test it would otherwise pass.
How to check
Before a retest, drive at least fifteen to twenty minutes including sustained highway speed, and go straight to the station without a long wait. Verify the coolant temperature reading has reached its normal range on a scan tool. If the first failure was marginal and the car was cold, retest properly warmed before spending money.

Ordering reflects how often each cause is responsible in general, not a probability for your vehicle. Confirm by testing before replacing anything.

Codes commonly stored with this symptom

If the light is on

A stored trouble code narrows this considerably — it tells you what a module actually measured rather than what you noticed. Any basic scanner will read it. Look the code up here.

Does the report say you failed for incomplete readiness monitors rather than for a code or a gas reading?

Was the check engine light on before the test?

Was the failure a catalyst efficiency code, P0420 or P0430?

If it was a tailpipe test, which reading was out of range — HC, CO, or NOx?

Was the engine fully warmed up when it was tested?

Has the vehicle had an aftermarket exhaust, intake, tune, or deleted emissions component installed?

Readiness monitors incomplete after a recent battery disconnect or code clear

Malfunction indicator lamp commanded on, or a stored emissions code

Catalyst efficiency below threshold

Rich or lean fuel mixture

Misfire

Evaporative emissions leak or gas cap fault

High oxides of nitrogen

Visual inspection or tampering failure

Engine tested cold or under invalid conditions