Documenting Intermittent Concerns on the Repair Order

Intermittent concerns are some of the most frustrating jobs in the shop. The customer swears the car does something weird, but when you drive it, everything works perfectly. You cannot fix what you cannot reproduce — but your documentation on that repair order is the difference between a protected shop and a liability nightmare. Here is how to handle intermittent complaints the right way, from the initial write-up through the follow-up visit.

Written by Anthony Calhoun, ASE Master Tech A1-A8

Why Documentation Matters More Than the Fix

On a normal repair, the fix speaks for itself. You replace the part, the symptom goes away, everybody is happy. Intermittent concerns are different. When you cannot reproduce the problem, your repair order becomes your only proof that you did your job. It protects you in three critical ways:

• Warranty audits: If this is a warranty claim, the manufacturer will review your documentation. A vague write-up like "checked vehicle, unable to duplicate" will get the claim kicked back. A detailed write-up showing exactly what you tested and what data you collected will hold up.
• Liability protection: If the customer has an accident and claims the intermittent concern caused it, your documentation is your defense. Courts look at what the technician documented, not what the technician remembers.
• Comeback prevention: When the car comes back — and it will — the next technician (or you, three weeks later) needs to know exactly what was already checked. Good documentation prevents duplicate work and wasted hours.

What to Document When You Cannot Reproduce the Fault

Every intermittent concern write-up needs five elements. Miss any one of them and your documentation has a hole.

1. The Customer's Exact Description

Write down what the customer said, not your interpretation. There is a big difference between "customer states vehicle shakes" and "customer states steering wheel vibrates side to side at 65 mph on the highway, lasts about 10 seconds, happens 2-3 times per week, started about a month ago." The second version gives you diagnostic direction. The first one gives you nothing.

2. Conditions They Reported

When does it happen? Cold start only? After driving 20 minutes? Only on the highway? Only when turning? Only when it rains? Get specific. These conditions tell you what system is involved and when to look for the fault.

3. What You Did to Attempt Reproduction

Document your test drive: how long, what speeds, what conditions you tried to replicate. "Test drove vehicle 12 miles including highway and city driving, cold start and at operating temperature. Attempted hard acceleration, deceleration, and steady cruise at 65 mph. Could not reproduce described vibration." That tells anyone reading it that you actually tried.

4. What You Found

Even if you could not reproduce the symptom, you probably found something. Scan results, history codes, pending codes, freeze frame data, fluid condition, visual inspection findings — all of it goes on the RO. "No codes stored" is still a finding. "All four tires at 35 PSI, tread depth even, no visible wheel damage" is still a finding.

5. What You Did Not Find

This is the one most techs skip, and it is just as important. "No DTCs stored, no pending codes, no history codes. All relevant PIDs within normal range during 15-minute test drive under described conditions. No abnormal noises, vibrations, or drivability concerns observed." That sentence tells the next tech — and the warranty auditor — that you checked and it was clean.

Using Scan Tool Data on Intermittent Concerns

Your scan tool is your best friend on intermittent complaints, even when the car is not acting up. Here is what to pull and document:

History and Pending Codes

Do not just check current codes. Pull the full code history — stored, pending, and permanent. A history code that matches the customer complaint is gold. It proves the system detected an event even if you cannot reproduce it. Always include the code number, description, and status (current, pending, or history) in your documentation.

Freeze Frame Data

Freeze frame captures the operating conditions when a code set. Engine RPM, vehicle speed, coolant temperature, fuel trim values, engine load — this snapshot often tells you more than the code itself. A P0301 that set at 2,200 RPM, 45 mph, and 195 degrees is a very different problem than a P0301 that set at idle, cold start, 72 degrees. Document the full freeze frame.

Mode $06 Data

Mode $06 shows test results for emission-related monitors. A component can be passing but trending toward failure. If the customer reports an intermittent misfire and Mode $06 shows cylinder 3 misfire counts at 80% of the threshold, you have found your problem — it just has not failed hard enough to set a code yet. This is incredibly valuable data on intermittent concerns.

Setting Up the Customer for a Productive Return Visit

When you send a car back without reproducing the concern, you are not done. You need to set the customer up to help you on the next visit. Tell them exactly what to note when it happens again:

• Speed: Approximately how fast were they going?
• Temperature: Cold start or warmed up? How long had they been driving?
• Road conditions: Highway, city, bumpy road, turning?
• Duration: How long did it last? Seconds or minutes?
• Warning lights: Was the check engine light on? Flashing? Any other warning lights?
• Weather: Raining, humid, cold morning, hot afternoon?
• Frequency: Every day? Once a week? Getting more frequent?

Some shops print a simple log sheet for customers with intermittent concerns. Hand it to them with the keys and say, "Next time it acts up, fill this out right away while it is fresh in your mind." This turns an unreliable narrator into a useful data source.

The Proper NTF (No Trouble Found) Write-Up

An NTF does not mean "nothing happened." It means you could not reproduce the concern during this visit. A proper NTF write-up includes:

• Customer's reported concern (verbatim)
• Conditions the customer described
• All tests performed (with specifics — mileage driven, time spent, conditions attempted)
• All scan data collected (codes, freeze frame, Mode $06, PIDs)
• Visual inspection results
• Your professional assessment: "Based on data collected, no fault detected at this time. Recommend customer document conditions when symptom recurs and return for further diagnosis."

Never write "could not duplicate" by itself. That tells nobody anything. An NTF should read like a lab report — here is what we were looking for, here is how we looked, here is what we found (or did not find), and here is what we recommend.

Data Loggers and Continuous Monitoring

For truly elusive intermittent concerns, consider setting up continuous monitoring. This is especially useful for faults that happen once a week or less.

OBD-II Data Loggers

A plug-in OBD-II logger can record PID data continuously while the customer drives. When the symptom occurs, you have a data trail leading up to and through the event. Look for anomalies in fuel trims, misfire counters, sensor readings, or module communication right before the customer-reported event.

Dash Cameras with OBD Integration

Some dash cameras record vehicle data alongside video. When the customer reports the event, you can correlate what the car was doing (speed, RPM, throttle position) with what it looked like from inside the cabin.

Manufacturer-Specific Flight Recorders

Some OEM scan tools have a flight recorder function that continuously logs data in a rolling buffer. When the event occurs, the buffer captures the data from before, during, and after the fault. GM's GDS2, Ford's IDS/FDRS, and Toyota's Techstream all have versions of this feature.

Real-World Examples

The Intermittent Misfire

Customer reports a stumble at idle, happens a few times a week, no check engine light. Scan shows no current codes, but Mode $06 shows cylinder 4 with 45 misfire counts (threshold is 60). Freeze frame from a history P0304 set three weeks ago shows idle, cold start, 85 degrees coolant temp. Diagnosis: ignition coil breaking down when cold. Swapped coil from cylinder 4 to cylinder 2 — misfire followed the coil. Replace coil, problem solved. Without Mode $06 data and the history freeze frame, this would have been an NTF.

The Random Stall

Customer reports engine stalls randomly in parking lots and drive-throughs. Never happens on the highway. No codes, no history codes. Installed an OBD-II data logger for one week. On day 4, the logger caught a fuel pressure drop from 58 PSI to 22 PSI over 8 seconds right before a stall event — at idle, in gear, AC on. Diagnosis: fuel pump failing under combined electrical load at low RPM. Traditional diagnosis would have missed this because the pump tests fine when the engine is running at normal load.

The Occasional ABS Activation

Customer reports ABS activates on dry pavement at low speed when braking normally. No ABS codes. Scan data shows all four wheel speed sensors reading correctly during road test. On the third visit, tech captures the event on a data recording — left front wheel speed signal drops to zero for 200 milliseconds during braking, triggering ABS intervention. Diagnosis: cracked tone ring on left front CV axle. The crack only opens up under braking load, causing a momentary signal loss. Visual inspection with the wheel off confirmed the hairline crack.

Communicating with Service Advisors

The advisor is your translator between the customer and the repair order. For intermittent concerns, make sure the advisor understands:

• You could not reproduce the problem, but that does not mean the customer is wrong
• What you did find (data, observations, trends)
• What you recommend for the next visit
• How to explain to the customer what happens next — without making promises you cannot keep

The worst thing an advisor can tell a customer on an intermittent concern is "we fixed it" when you actually wrote NTF. The correct message is: "We performed a thorough inspection and data analysis. We did not find a fault at this time, but here is what we need you to do if it happens again." That sets honest expectations and prevents a frustrated comeback.

When to Recommend Extended Testing

Some intermittent concerns justify keeping the vehicle longer. Recommend extended testing when:

• The concern is safety-related (stalling, loss of braking, loss of steering)
• The customer reports it happens daily but you cannot reproduce it on a single test drive
• You found data suggesting a developing fault (Mode $06 trending, history codes matching complaint)
• The vehicle is still under warranty and the manufacturer requires documented reproduction

Get authorization in writing before keeping a car overnight or for extended testing. Document every day — what was tested, how long, what conditions, what was found. Even if nothing happens on day one, that documentation proves you were actively working the problem.