---
title: "Symptoms of a Bad Alternator: What 1,231 Owner Reports Describe"
description: "Not another symptom list. We counted 1,231 owner reports filed with NHTSA that name the alternator: what they describe, in what order, and where the car was."
url: "https://baronauto1.com/car-care/symptoms-of-a-bad-alternator/"
type: "article"
published: "2026-09-07"
modified: "2026-09-07"
site: "Baron Auto"
disclaimer: "This site is under new ownership and is not affiliated with Baron Auto Emporium dealership."
---

# Symptoms of a Bad Alternator: What 1,231 Owner Reports Describe

> Not another symptom list. We counted 1,231 owner reports filed with NHTSA that name the alternator: what they describe, in what order, and where the car was.

*Detailing & Maintenance · 25 min read · 5,680 words*

## The short version

- This page is counted, not remembered. We pulled every owner complaint the federal government holds for 32 nameplates across six model years — 130,243 distinct records — and read the 1,231 of them whose owner wrote the word alternator.
- The symptom owners describe most is not a dim headlight. It is the engine stopping. Stalling, shutting down or losing power appears in 504 of the 1,231 accounts; dimming or flickering lights, the symptom every list leads with, appears in 60.
- There is no reliable warning sequence, and that is the finding that matters. Among accounts that mention both a dashboard warning light and a stall, the light is named first exactly half the time. You cannot wait for the next symptom, because there is no next symptom.
- The one thing that does come last is the tow truck. Of the accounts mentioning both a stall and a tow, 160 of 174 put the stall first.
- It is mostly a roadside failure, not a driveway failure. Where the account says where the car was, 503 say it was moving, against 106 parked.
- It is a summer fault. June to September holds 42.2% of these accounts against 36.5% of the whole complaint file over the same months — a real gap, because the whole file is its own control.
- The battery gets replaced first. Of the 170 accounts that mention replacing a battery, 143 name the battery before they name the alternator. On a used car for sale, a new battery beside an untouched charging system is a question, not a feature.
- What the file cannot tell you: how long an alternator lasts. Vehicle age at failure in these accounts is identical to vehicle age across the whole complaint file, so age separates nothing at all.

Search this question and you get the same nine bullets everywhere: dim headlights, a battery warning light, a whining noise, a burning smell, electrical gremlins, a dead battery, difficulty starting, stalling, a check-engine light. The list is not wrong. It is simply written from memory, copied sideways, and ordered by nothing at all — which is why it cannot answer the question a driver is actually asking, namely: is what is happening to my car right now the charging system, and how long have I got?

There is a public record that can do better. The National Highway Traffic Safety Administration takes complaints from vehicle owners and publishes them, free text and all, through an open interface. The narratives are not diagnoses and they were not written for us. They are what people said happened, in their own words, mostly after somebody had told them what the part was. That makes them a poor guide to how often alternators fail and an unusually good guide to what an alternator failure looks like from the driver’s seat.

So we pulled a large sample of them, archived the raw responses, and counted. What follows is what the counting produced — including one result that contradicts how almost every symptom list on the internet is organised, and two places where the data has nothing to say and we say so instead of guessing.

## What 1,231 owner accounts actually describe

The pull covers 32 high-volume nameplates across six model years — 2008, 2011, 2013, 2015, 2017 and 2019 — chosen for spread across make and body style rather than for what they might return. That is 192 responses and 130,243 distinct complaints once duplicates are removed.

The first thing you learn is that the federal database has no alternator. There is no such component heading. The nearest is ELECTRICAL SYSTEM, which holds 22,064 of those records and also carries window switches, wiring harnesses, body control modules and infotainment faults. It is far too wide a bucket to answer a question about one part.

It is also, awkwardly, too narrow. Filter instead on the narrative — every account whose owner wrote the word alternator — and you get 1,231 records. Only 943 of them, 76.6%, are filed under ELECTRICAL SYSTEM at all. The other 288 sit under headings the owner or the intake handler thought fitted better. Fewer than half of the whole set, 556 of them, carry ELECTRICAL SYSTEM as their only heading. If you had searched by component you would have missed roughly a quarter of the file.

Two limits before any number below is read as more than it is. First, filtering on the word selects for owners who had already been told what the part was, usually because somebody had diagnosed or replaced it. The driver who saw a light, never found out why, and got on with life is not in this file. So the counts under-represent the early, ambiguous stage of the fault. Second, these are shares of accounts that name the alternator, never shares of alternator failures in the world. Nothing here is a failure rate, and we will not present it as one.

With that said, here is what people describe. Stalling, shutting off, cutting out or losing power appears in 504 of the 1,231 accounts — 40.9%, more than any other symptom by a wide margin. A tow appears in 302, or 24.5%. A dashboard warning light of some kind appears in 294, or 23.9%. A refusal to start, or a battery found flat, appears in 240, or 19.5%. A burning smell or visible smoke appears in 175, or 14.2%. A radio, screen or navigation unit going dead appears in 144, or 11.7%. Steering going heavy appears in 139, or 11.3%.

And dimming or flickering lights — the symptom that leads every list, the one that gives the failure its folklore — appears in 60 accounts. That is 4.9%. It is the second-rarest thing in the set. The rarest is a loss of brake assist, which turns up in 13 accounts, or 1.1%, too few to build any argument on and reported here only because leaving it out would be the kind of tidying that makes data lie.

None of that means a dimming headlight is not a charging fault. It means that by the time owners are moved to write to a federal agency, the thing they are writing about is usually not a headlight. It is a car that stopped.

Which raises the obvious next question, and the one every symptom list quietly implies it has answered: in what order do these things arrive?

**Figure: Which symptom the account names first**

A horizontal bar chart of nine symptom pairs, each drawn only from the alternator accounts that mention both, showing how often the first-named symptom leads and how many accounts are behind each split. Seven pairs sit at or close to an even split and only the two involving a tow break away. The two thinnest pairs, on 14 and 20 accounts, are drawn faintly.

This measures the order owners tell it in, not the order the events happened — people summarise, backtrack and lead with whatever frightened them most. Read that way it still settles something: there is no queue. Of the 138 accounts mentioning both a dashboard warning light and a stall, the light is named first in 69 and the stall in 69. The only dependable ordering in the file is the tow, which comes last in 160 of the 174 accounts describing both. The pairs resting on 14 and 20 accounts are drawn faintly because fourteen accounts is fourteen accounts. These are shares of accounts naming both items, out of the 1,231 whose owner wrote the word alternator; filtering on the word selects for owners who had already been told what the part was, so nothing here is a failure rate or a comparison between makes.

## The symptoms do not arrive in a reliable order

Symptom lists are written as ladders. First the light, then the dimming, then the stall, then the no-start. The implication is a schedule: notice the early rung and you have time; reach the late rung and you do not.

You can test that on this file, because many accounts mention more than one symptom, and text has an order. For every pair of symptoms, take only the accounts that mention both, and count how often the first one is named before the second. That measures the order owners tell the story in, which is not the same as the order events happened — people summarise, backtrack and lead with whatever frightened them most. But if there were a strong underlying sequence, it would show through the noise. Across nine pairs, mostly, it does not.

The clearest case is the pairing every list treats as certain. Of the 138 accounts that mention both a dashboard warning light and a stall, the light is named first in 69 and the stall in 69. That is an exact split. Take the warning light against a refusal to start: 49 accounts mention both, and the light comes first in 28 of them and second in 21 — 57.1%, which on 49 accounts is not a sequence, it is a coin. Dimming lights against a no-start: 14 accounts, 7 each way, 50%. Dimming lights against a warning light: 20 accounts, 11 and 9, 55%. A dead radio or display against a no-start: 31 accounts, 17 and 14, 54.8%.

Two pairs lean, mildly. A warning light is named before a burning smell in 33 of the 53 accounts mentioning both, 62.3%. A stall is named before a no-start in 61 of 89, 68.5%, which makes sense — a car that dies while running is a car that then will not restart.

And two pairs are not close at all, because they are not really about symptoms. Of the 87 accounts mentioning both a warning light and a tow, the light comes first in 73, or 83.9%. Of the 174 mentioning both a stall and a tow, the stall comes first in 160, or 92%. The tow truck is the end of the story in almost every account that has one. That is the only dependable ordering in the file.

The small samples on some of those pairs are real and we are not going to dress them up: fourteen accounts is fourteen accounts. But the pattern holds across pairs of every size, including the 138-account pair, and it points one way. There is no queue. A driver does not get a warning light, then a grace period, then dimming, then a stall. Sometimes the light is first. Sometimes the car simply stops and the light is something the owner noticed afterwards, or mentioned second because the stall was the point.

The practical consequence is the most useful sentence on this page. **You cannot wait for the next symptom, because there is no next symptom.** Whatever you have got is not necessarily an early rung on a ladder, and treating it as one is how people end up on a hard shoulder. If the charging system is under suspicion, the question is not how many more symptoms you have left before it is serious. It is where you would rather be standing when it stops.

## Where the car was when it happened

A flat battery is a driveway event. You come out in the morning, the car is dead, you are annoyed and late. A great deal of charging-system advice is written as though a failing alternator were the same kind of inconvenience, only more expensive.

The accounts do not read that way. Classify each one by whichever situation it states first, and 503 of the 1,231 describe a car that was moving — driving, on the highway, in traffic, in the middle of an intersection. Only 106 describe a car that was parked or standing, and 136 describe somebody trying to start it. The remaining 486, very nearly two in five, never say where the car was at all, which is worth stating plainly rather than distributing among the others.

Counted a different way — any mention anywhere in the account rather than the first one — 538 accounts, 43.7%, mention driving; 197, 16%, mention a failed start; 165, 13.4%, mention the car parked. The ranking does not move.

So the modal experience in this file is not a dead car in a driveway. It is a car that quits while it is being driven, and that is a materially different event, because everything the driver would normally use to deal with it — the steering assistance, the brake assistance, the lights, the ability to restart — is downstream of the thing that has just failed.

The agency’s own severity flags on these 1,231 accounts: 86 are flagged as involving fire, 7% of the set; 14 are flagged as involving a crash; 14 injuries are recorded across the whole set, and no deaths. Those flags are set at intake from what the owner reported, they are not investigation findings, and a fire flag on an account mentioning an alternator does not establish that the alternator caused the fire. But the flags exist on this component’s accounts at a rate that argues against treating a suspected charging fault as an errand to be run next week.

## What goes down with it

The reason a charging failure is different from a battery failure is architectural. A battery’s job finishes the moment the engine catches. From then on the alternator runs every electrical load on the car and refills the battery besides, so when it stops supplying, the car does not lose one system. It starts spending down a reserve that everything else is drawing on at once.

The filing pattern shows that better than any description could. Because a complaint can carry several component headings, you can see what else the owner or the intake handler thought had failed at the same time. Alongside the 943 filed under ELECTRICAL SYSTEM, 313 of the 1,231 accounts — 25.4% — also carry ENGINE. 116 carry UNKNOWN OR OTHER. 87 carry STEERING and another 87 carry POWER TRAIN, both 7.1%. 69 carry ELECTRONIC STABILITY CONTROL, 5.6%. 66 carry FUEL/PROPULSION SYSTEM, 5.4%. 65 carry SERVICE BRAKES, 5.3%. 46 carry EXTERIOR LIGHTING, 3.7%, and 35 carry VEHICLE SPEED CONTROL, 2.8%.

Read that list as an argument rather than a table and it says something specific about modern cars. Steering assistance on most of them is electric. Stability control is electronic. Speed control is electronic. Fuel delivery and ignition are electronic. When supply fails, those are not separate faults that happened to coincide — they are one fault, reported through whichever system the driver noticed losing first. That is also why the component heading was never going to isolate this part, and why 288 of these accounts are filed somewhere other than the electrical system.

## The three faults this gets confused with

Nearly every symptom in the list above is produced by at least one other failure, and two of those failures are cheaper to fix than a charging system. That is exactly why so many of these accounts describe money spent on the wrong part first. The table below is the discrimination, not the diagnosis: what separates the charging system from the battery, from the starter, and from a drive belt that has stopped turning it.

| What you notice | Also produced by | What separates them |
| --- | --- | --- |
| Dead in the morning, fine after a jump | A tired battery, or something drawing current while the car is parked | This one points away from the charging system. If the car runs normally once started and only dies after standing, the alternator was doing its job while the engine was running. |
| Starts on a jump, then dies within a minute or two | Very little else | This is the charging system almost by elimination. The battery had enough to crank; nothing replaced what it spent. A car that will not stay running on its own is not a battery problem. |
| Battery warning light while driving | A broken or slipping drive belt, a wiring or connection fault, a control-module fault | The light is named for the battery and wired to the charging circuit. It reports that supply has dropped, not that the battery is flat. What it does not do is tell you which link in the chain dropped it. |
| Slow, laboured cranking | A tired battery, corroded battery terminals or cables, a failing starter | Cranking is the battery’s job. A charging fault reaches cranking only at the end, after several drives have drained the battery. Slow cranking that got worse over a cold week is usually the battery. |
| One loud click, or nothing, and the lights stay bright | A failed starter or its solenoid | Bright lights mean the battery has charge. A charging failure that has reached no-start leaves the car dim and dead, not bright and silent. |
| Engine stops while driving | Fuel delivery, ignition, a crank or cam sensor, a control module | Whether anything electrical survived. A charging-system stall takes the dash, the radio and the assistance with it. A fuel or ignition stall usually leaves the electrics working while the engine will not run. |
| Steering suddenly heavy, or brakes suddenly hard | A failed steering assist unit, a vacuum or booster fault, a snapped drive belt | Whether it happened alone or with company. Losing assistance on its own points at that system. Losing it in the same second as the dash lighting up points upstream at supply. |
| Second flat battery within a year or two | A parasitic drain, short trips that never fully recharge, a genuinely faulty battery | Whether anyone tested the charging system before fitting the second battery. This is the sequence most often described in the file, and it is the one most often solved by buying the wrong part twice. |

One symptom is deliberately absent from that table. Noise — a whine, a growl, a grinding bearing — is a fixture of every published list and is barely present in these narratives. That is probably a fact about what people write to a safety regulator rather than a fact about alternators, but we are not going to print a count we do not have. If your car has developed a bearing noise, the file has nothing to say about it and neither will we.

## The pattern in the file: the battery gets replaced first

170 of the 1,231 accounts — 13.8% — mention a battery being replaced or changed. In 143 of those 170, the owner names the battery before they name the alternator. In only 27 does the alternator come first.

That ordering is worth more than it looks, because these are the same narratives whose symptom orderings were coin flips. The symptoms do not sequence, but the spending does. The battery is the cheaper part, the more familiar part, the part that any parts counter will test and sell you in fifteen minutes, and it is the part that gets bought first — and then, some weeks or months later, the account continues and the alternator finally gets named.

62 accounts, 5%, go further and mention a second, another or a new alternator, or count alternators in the plural. A charging system that has been replaced and failed again is a different problem from one that failed once: a belt, a wiring fault, an over-loaded electrical system or a poor part can all send a fresh unit back the same way, and none of them is fixed by fitting a third.

For anyone buying rather than repairing, this is the single most useful thing on the page. A recent battery sitting beside an original charging system, on a car that is being sold, is exactly what the front half of these 143 accounts looks like. It is not proof of anything. It is one question, asked once, that costs nothing: *why was the battery replaced?* A straight answer about age or a failed test is fine. Vagueness about a car that kept going flat is not. Our page on [what actually drives a battery bill](https://baronauto1.com/car-care/car-battery-replacement-cost/) covers the same crossing point from the battery’s side, and [what decides an alternator bill](https://baronauto1.com/car-care/alternator-replacement-cost/) covers why we do not print a figure for either.

## It is a summer failure, not a winter one

Batteries are a winter story, and the association is so strong that charging faults get filed under it by association. The file disagrees, and it disagrees against a control, which is the only way a seasonal claim is worth anything: complaint volume has its own shape across the year, so an alternator month share means nothing until you put the whole file’s month share beside it.

**Figure: The alternator year against the whole complaint file**

A grouped bar chart setting the share of alternator accounts falling in each window or month against the same share across all 130,243 complaints in the archive. The first two rows are four-month windows and the remaining four are single months, so each pair is read within its own row rather than across rows.

The control is the figure. Complaint volume has its own shape across the year, so an alternator month share means nothing until the whole file’s share is put beside it. June to September holds 42.2 per cent of the alternator accounts against 36.5 per cent of the archive; December to March holds 27.9 against 32.2, and the gap runs the same way in both windows. August is the peak at 11.6 against 9.1, July runs 11 against 9.8, September 10.3 against 8.1, and February is the trough at 6 against 7.6. Read each pair within its own row and never across rows — the first two are four-month windows and the rest are single months. The mechanism is not in the data and none is claimed here. These are complaints owners chose to file, not failure rates for any part or any make.

June to September holds 42.2% of the alternator accounts and 36.5% of all 130,243 complaints in the archive. December to March holds 27.9% of the alternator accounts against 32.2% of the whole file. The gap runs the same way in both windows and it is not small. Month by month, August is the peak: 11.6% of alternator accounts against 9.1% of the file. July runs 11% against 9.8%, September 10.3% against 8.1%. February is the trough at 6% against 7.6%.

We are not going to tell you the mechanism, because the file does not contain one. Under-bonnet heat, air-conditioning load, longer trips and more driving in summer are all plausible and none of them is measured here. What the data supports is the pattern and the direction, nothing further.

What it changes for a reader is modest but real: a car that has been flawless all winter has not proved anything about its charging system, and a battery that struggled in January is not evidence that January was the problem. If you are buying in spring, you are buying ahead of the season this fault shows up in.

## Where in a car’s life it lands, and where the file goes silent

A minority of these accounts carry an odometer reading in a fixed sentence written by the intake handler rather than by the owner. That is the only mileage in the record that can be read without inventing an interpretation, and it appears in 216 of the 1,231.

On those 216, the middle reading is 75,000 miles. A quarter are below 49,600 and a quarter above 111,785. 147 of them — 68.1% — are below 100,000, and 79 — 36.6% — are below 60,000.

Two comparisons make those numbers mean something. Against the ELECTRICAL SYSTEM heading as a whole, whose 3,272 readings have a midpoint of 65,000, alternator accounts arrive later: this is not one of the faults that surfaces in the first few years alongside switches and modules. Against the whole complaint file, whose 23,560 readings have a midpoint of 75,123, they arrive at the same time as everything else. An alternator, in this file, is an ordinary mid-life event.

And now the silence, which is the honest half of the section. **None of this is an alternator’s life expectancy.** These are the mileages at which people complained to a regulator, not the mileages at which charging systems fail, and the two differ by every failure that was quietly fixed and never reported. The readings are also right-censored: these vehicles are still being driven, so failures that have not happened yet cannot be in the file.

Vehicle age says even less. We measured it expecting something and found nothing: the alternator accounts sit at a first quartile of 3 years, a midpoint of 5 and a third quartile of 7 — the identical quartiles to all 130,243 complaints in the file. Age at failure does not distinguish this fault from anything else, and where a measurement comes back flat the right thing to do is say so rather than find a way to phrase it as a finding.

## What to do when you suspect it

This is not a repair page and there is nothing below about getting under a car. It is about the decision, which is the part people get wrong.

Start from what this page has established. There is no sequence, so there is no grace period to plan around. The failure is mostly a moving failure, so the risk is not being late for work. And the car is running on a reserve that is not being refilled, which means every minute of driving, every headlight, every wiper, every seat heater is spending down a battery that has to hold enough to keep the engine running at all.

The practical translation: when the battery warning light comes on while driving, the job is to stop somewhere you would be content to be stopped. That is a decision about geography, not about mechanics. Somewhere lit, off the carriageway, reachable by a recovery truck. It is not a decision to defer to the next exit on the grounds that the car still feels fine. The only dependable ordering in this whole file is that the tow comes last — in 160 of the 174 accounts describing both, the stall precedes it — and a car that still feels fine is a car that has not got there yet.

Turning off what you can — air conditioning, heated glass, anything drawing current that you do not need to keep moving safely — buys some of the reserve back. It does not fix anything and it should not be read as permission to keep going.

A jump start deserves its own caution here, because it is the thing people reach for and it is diagnostic if you watch what happens next. A car that starts on a jump and then dies again shortly after the leads come off has told you something the charging system did not survive. Our guide to [jump starting a car properly](https://baronauto1.com/car-care/how-to-jump-start-a-car/) covers doing it without damaging anything; what it will not do is get you home on a failed alternator, and attempting a long drive on a jump-started car with no charging is how a roadside problem becomes a motorway one.

The test itself belongs to somebody with the right equipment, and it is quick. A charging system is measured, not guessed at, and a workshop or a parts counter that tests both the battery and the charging output will separate the two in minutes. What matters is that both get tested. The 143 accounts in this file where the battery was named first are, in aggregate, a very expensive argument for insisting on that.

If the car will not start at all and you are trying to work out where in the chain the problem sits, our page on [a car that will not start](https://baronauto1.com/car-care/car-wont-start/) walks the whole decision tree, of which the charging system is one branch.

## On a used car you are about to buy

Everything above is written for somebody who owns the car. The buyer’s version is shorter and more useful, because a buyer has leverage a current owner does not.

Look for the sequence, not the symptom. A charging fault on a car for sale rarely announces itself on a twenty-minute test drive — a partially failing alternator can hold up fine on a short run with a healthy battery. What it does leave behind is a paper trail and a parts trail: a battery that is obviously newer than the car and newer than everything around it, a receipt for a battery and nothing else, or an owner who mentions the car “going flat a couple of times” and then changes the subject.

Ask the age of the battery and why it was replaced. Ask whether the charging system was tested at the same time. Both are ordinary questions with ordinary answers, and the answer you are listening for is whether anyone ever established a cause.

Then, on a car you are serious about, have the charging output measured as part of a pre-purchase inspection rather than trusting a dashboard that only lights up once supply has already dropped. Our guide to [what a pre-purchase inspection should cover](https://baronauto1.com/buying-guides/used-car-pre-purchase-inspection/) sets out how to brief an inspector; a charging test takes them a couple of minutes and it is worth naming explicitly when you book.

One last framing, because it is the thing this whole page has been circling. The alternator is not an expensive part in the scheme of a used car, and this page is not an argument for walking away from a car that needs one. It is an argument for knowing which part you are buying. The costly version of this fault is not the repair. It is the battery first and the alternator second — the order 143 of these accounts name them in — with a tow somewhere in the middle. One test at the start separates the two parts, and it is the cheapest thing in the sequence.

## Common questions

### How long can I drive with a failing alternator?

Nobody can give you a number, and anyone who does is guessing. It depends on how much output is left, how charged the battery was when the fault started, and how much current the car is drawing. What this file can tell you is that a tow appears in 302 of the 1,231 accounts, and that where both a stall and a tow are described, the stall comes first in 160 of 174. The realistic planning assumption is not a distance. It is that the car will stop, at a time you do not choose.

### Will a new battery fix it?

If the charging system has failed, no — it will buy you the length of one battery charge and then leave you in the same place, poorer. This is the most common expensive mistake in the file: 143 of the 170 accounts that mention a battery replacement name the battery before the alternator. The reverse is also true and matters just as much: if the battery is the actual fault, replacing a healthy alternator fixes nothing either. That is why both get tested, together, before either is bought.

### Does a dimming headlight always mean the alternator?

No, and in this file it is not even the common presentation. Dimming or flickering appears in 60 of the 1,231 accounts, 4.9%, the second-rarest symptom counted. Lights that dim and brighten with engine speed are a genuine charging-system tell and worth acting on. But a car whose headlights have simply grown yellow and weak over years is describing bulbs and reflectors, not supply.

### Can a failing alternator make the engine stall?

It is the single most described symptom in the file: 504 of 1,231 accounts, 40.9%, describe stalling, shutting off, cutting out or losing power. On a modern engine, ignition and fuel delivery are electrically operated, so when supply drops far enough the engine stops running. That is also why a charging-system stall usually takes the dashboard and the radio with it, which is the clearest way to distinguish it from a fuel or ignition stall that leaves the electrics alive.

### Why did it fail in summer? I thought this was a winter problem.

The winter association belongs to batteries, and it is well founded. It does not transfer. In this archive, June to September holds 42.2% of the alternator accounts against 36.5% of all complaints in the same months, and December to March holds 27.9% against 32.2%. August alone runs 11.6% against 9.1%. The mechanism is not in the data and we will not invent one, but the pattern is clear enough that a trouble-free winter proves nothing about a charging system.

### The dashboard light has not come on. Does that rule it out?

It does not. A dashboard warning light of any kind appears in only 294 of the 1,231 accounts, 23.9% — and where accounts mention both a light and a stall, the light is named first in exactly half of the 138 cases. The light reports that supply has already dropped below a threshold. It is not an early warning system and this file gives no support for treating it as one.

### Is a second alternator in a few years normal?

It is not common in this file — 62 accounts, 5%, mention a second, another or a new alternator, or count them in the plural — and it usually means the first replacement treated a symptom. A belt that is slipping, a wiring or connection fault, a battery that was never any good, or an electrical load the system was not designed for will all shorten a new unit’s life the same way they shortened the old one. Two failures is a reason to look upstream, not a reason to fit a third.

### Why does this page not tell me what the repair costs?

Because a figure would be the least reliable sentence on the page and the least useful. What sets the bill is which part gets fitted, the output the vehicle specifies, how much has to come off the engine to reach it, whether the belt and its hardware are replaced alongside it, and whether the car has to be told about the new part. Those vary more between two cars in the same car park than any average would suggest. [Our page on what decides an alternator bill](https://baronauto1.com/car-care/alternator-replacement-cost/) takes each of those in turn, and it prints no figure either, for the same reason.

### Where does this data come from, and can I check it?

NHTSA’s complaints interface, which is public and free. We requested every complaint for 32 nameplates across 2008, 2011, 2013, 2015, 2017 and 2019, archived the 192 raw responses with their retrieval date, deduplicated to 130,243 records, and read the 1,231 whose narrative names the alternator. Every count on this page is regenerated from that archive by a script rather than typed in, and the counts are shares of accounts, not rates of failure. The two places the data came back empty — vehicle age, and any usable count for bearing noise — are reported as empty above rather than filled in.

## Sources and further reading

- [NHTSA Office of Defects Investigation complaint database](https://www.nhtsa.gov/nhtsa-datasets-and-apis)

Recall, complaint and safety-rating figures on this page were retrieved from the federal databases above on August 19, 2026. Federal data changes — re-check any VIN before you rely on it.

---

*This site is under new ownership and is not affiliated with Baron Auto Emporium dealership.*

Canonical source: https://baronauto1.com/car-care/symptoms-of-a-bad-alternator/