How Often Should You Change Spark Plugs?

Search this question and you will be told, repeatedly and confidently, that spark plugs should be replaced every 30,000 miles. Six separate franchise dealer service blogs on the first page of results state it as a flat fact. Not one of them cites a source, and it contradicts every manufacturer schedule we were able to source for any modern vehicle — including the schedules published by the same brands those dealerships sell.

The figure is not invented from nothing. It is roughly right for a conventional nickel-alloy plug of the kind that was standard equipment before the late 1990s, and it survives in circulation because it was once true. Applied to a car built in the last twenty years with long-life plugs from the factory, it is wrong by a factor of two to four, and it happens to be wrong in the direction that sells service work.

The table that dominates this topic — copper 20,000 to 30,000 miles, platinum 60,000 to 100,000, iridium 100,000 to 120,000 — appears in Google's AI summary, on Kelley Blue Book, on AutoZone and on dozens of dealer blogs. It does not appear on the website of any spark plug manufacturer we could find. We went looking for the primary source and there is not one.

NGK, which is one of the two largest plug makers in the world, publishes no material-based mileage at all. Its resource page on this exact question says only that the best guide is the manufacturer's recommendation for your vehicle, because the service varies from brand to brand and model to model. A company whose revenue comes from selling replacement spark plugs is declining to tell you how often to replace them, which is a fairly strong signal about how knowable the answer is.

There is also a unit problem hiding in plain sight. DENSO's platinum figure is 100,000 kilometres. The claim circulating in US content is that platinum plugs last 100,000 miles. Those are not the same statement — the second is 61 percent larger than the first. We cannot demonstrate that the American figure descends from a mishandled conversion of the Japanese one, and we are not asserting that it does. But any figure you encounter for this topic is worth checking for its unit before you act on it.

We could not retrieve first-party service-life figures from Bosch, Champion or Autolite within this research, and the per-material table that surfaced under their names traces to a third-party reseller blog rather than to the manufacturers. We are not attributing any interval to them. The same applies to ruthenium plugs, for which we found no published figure at all.

Among the secondary sources that do publish numbers, the disagreement is not marginal. JD Power states that the service life of platinum and iridium plugs is not more than 55,000 miles. Kelley Blue Book gives 60,000 to 100,000 miles for the same materials. KBB's ceiling is nearly twice JD Power's floor, and both are mainstream automotive publishers writing for the same audience in the same market.

Neither is obviously wrong. They are describing different vehicle populations and, probably, different risk appetites — a conservative figure that will be safe across almost any car versus a range that reflects what factory schedules actually say on the long end. What no honest page can do is average 55,000 and 100,000 and print 77,500, which is a number nobody stated and no vehicle was engineered around.

Toyota publishes three different answers

The clearest illustration comes from a single manufacturer contradicting itself. Toyota's consumer-facing car-tips page says spark plugs usually last around 80,000 miles. Toyota's maintenance schedule for the 2021 Camry gives 120,000 miles for the 2.5L four-cylinder and 60,000 miles for the 3.5L V6 in the same car.

The 80,000-mile consumer figure matches neither engine in Toyota's own best-selling sedan. It sits between them, which is what a smoothed average does. We are quoting it as evidence of how this confusion propagates, not repeating it as guidance — and the same caution applies to every generalised figure on this page, including the ones we have published.

The best available evidence for this is not a technical paper — it is the manufacturers' own schedules, where the same company gives its boosted engine a much shorter interval than its naturally aspirated one. Hyundai and Kia specify 45,000 miles or 36 months for the 1.6L turbocharged GDI and 105,000 miles or 84 months for the 2.0L naturally aspirated GDI. That is a 2.3x difference within one lineup, with the same corporate engineering standards behind both numbers.

The mechanism is straightforward. Cylinder pressure at the moment of spark determines how much voltage it takes to ionise a path across the gap — the denser the mixture between the electrodes, the higher the breakdown voltage. A boosted engine is therefore already running closer to its ignition system's voltage ceiling when the plugs are new, so it has less headroom left as the gap erodes. NGK's own product material for boosted applications describes designs intended to limit the quenching effect and reduce misfires caused by high cylinder pressures blowing out the spark.

Direct injection compounds this from a second direction. Fuel is sprayed into the chamber at very high pressure during the compression stroke rather than mixed into the intake air, so the charge around the plug is less homogeneous and can be locally richer. DI engines are also more prone to combustion-chamber and injector-tip deposits, which is a related maintenance story rather than the same one.

This is the best-sourced part of the topic, because it comes from the plug manufacturers directly rather than from anyone quoting them. What fails in a spark plug is not the plug — it is the gap. Every spark erodes a small amount of metal from the centre and ground electrodes, and the distance between them slowly grows.

DENSO publishes a rate. For normal plugs, electrode wear is approximately in the range of 0.10 to 0.15 mm for each 10,000 km driven — roughly 0.16 to 0.24 mm per 10,000 miles. Against a typical new gap of about 1.0 to 1.1 mm, that means a conventional nickel plug can grow its gap by more than 20 percent in 60,000 miles. Iridium and platinum exist specifically to slow this down: the noble metals have far higher melting points and resist spark erosion, which is why a fine-wire iridium centre electrode holds its edge long after a nickel one has rounded off.

NGK explains why the gap is what matters, verbatim: as spark plugs grow older they lose their sharp edges as material is slowly eroded away, and as the gap between those two points grows, the voltage required to bridge the gap increases proportionately. Even the best ignition systems will be strained to supply enough voltage to completely burn the fuel.

Do worn plugs destroy ignition coils?

You will be told they do, usually in the sentence immediately before a quote. The physical chain is genuinely not in dispute and is published by first parties: the gap grows, the required voltage rises, and the coil has to deliver more secondary voltage to fire it. From there, coil manufacturers and the shop-facing trade press assert that the additional strain causes overheating and eventual coil failure, and that the higher voltage stresses the coil's internals and sometimes the ECU driver circuit as well.

What we could not find is any evidence for the failure claim itself — no controlled study, no manufacturer warranty-data analysis, no published comparison of coil failure rates between vehicles serviced on interval and vehicles run long. Every source asserting the causal link is either a coil vendor or trade press written for shops, which is to say every source has a commercial or professional reason to make the claim.

The standard symptom list for bad spark plugs contains six items presented as equally meaningful. They are not remotely equally meaningful, and sorting them is more useful than repeating them.

The last row deserves emphasis because it drives a lot of unnecessary work. Seasonal MPG variation from cold air, winter fuel blends and more idling is large enough on its own to account for what most owners notice and attribute to their plugs.

The code distinction that tells you plugs from coils

Worn plugs present as misfire codes, and which misfire code you have narrows the fault considerably — a distinction we did not find made on any competing page.

• P0300 — random or multiple cylinder misfire. Consistent with an entire worn set, because a set wears fairly evenly. This is the classic end-of-interval presentation and the strongest code-level match for plugs.
• P0301 and P0303 — single cylinder misfire. A misfire isolated to one cylinder points at that cylinder's plug, coil or injector, and is more often a failed coil than a worn plug, precisely because plugs wear together and coils fail individually.
• P0420 — catalyst efficiency below threshold. The downstream consequence rather than the cause. Sustained misfire dumps unburnt fuel into the converter and can destroy it, which is the expensive outcome of ignoring a P0300.
• P0171 — system too lean, bank 1. Listed here as a differential diagnosis, not a plug symptom. A lean condition produces misfire symptoms that feel identical to worn plugs, and telling the two apart before buying parts is worth the trouble.

For the driveability side of this, see car jerks when accelerating, which is the symptom a load-dependent misfire produces most characteristically, and car shakes at high speed if the vibration only appears at sustained speed. A cylinder that is not firing passes unburnt fuel out of the exhaust, which is covered under black smoke from exhaust.

Almost every page on this subject argues for on-time replacement using performance and fuel economy. Those are the weakest arguments available. The strong one is mechanical and almost nobody makes it: past a certain point, the question stops being whether the plug still sparks and becomes whether it can be removed at all.

Nearly all modern cylinder heads are aluminium and nearly all spark plugs are steel. Two dissimilar metals in contact, cycled through several hundred degrees thousands of times, seize progressively. The longer the plug stays in, the more thoroughly it welds itself in place, and removal at that point can pull the threads out of the soft aluminium — a repair that means a thread insert, sometimes with the head still on the car and sometimes not.

Ford's Triton three-valve V8s

The best-documented case is Ford's 4.6L, 5.4L and 6.8L three-valve V8s, which used a two-piece Motorcraft plug with an extended shell prone to carbon-welding itself into the head. Ford issued technical service bulletin 08-7-6 in 2008, an eleven-page document giving a specific removal procedure and covering three distinct ways the plug can break apart on the way out. Its central instruction, in capitals in the original:

A dedicated extraction tool for the broken portion of these plugs exists as an off-the-shelf retail product, which is about as clear an indication as you could ask for that this failure is routine rather than exceptional.

The general lesson holds well beyond Ford. A plug left to 150,000 miles in an aluminium head is a materially harder removal job than the same plug at 100,000, and the cost of getting it wrong is not the price of a plug. This is the reason to service on schedule even when the car feels perfectly fine, and it is a reason we did not find given on a single competing page.

For a plug still comfortably within its interval, essentially no. There is no mechanism by which a plug with a correct gap and intact electrodes produces a better spark than another plug with a correct gap and intact electrodes. Restoring lost performance requires that performance have been lost, and DENSO's own wear curve says the demand-voltage rise is steep only in the first few thousand kilometres and shallow thereafter.

Claims of fuel economy and horsepower gains from early replacement circulate very widely — 3 to 8 percent, up to 15 percent, worn plugs cutting economy by 30 percent. We could not trace a single one of them to a controlled test. Every source was a shop blog or a parts-seller page with no stated methodology, so we are not publishing any of those figures, not even as a range.

• Replacing genuinely worn or overdue plugs can restore lost performance and economy. This is defensible, because there was a loss to restore.
• Replacing in-spec plugs early buys you almost nothing except the elimination of seizure risk, which is real but modest if you are nowhere near the interval.
• Every removal is itself a risk. Each time a plug comes out of an aluminium head there is an opportunity to cross-thread or mis-torque it going back in. Doing this more often than necessary is not free.

The practical conclusion is unglamorous: follow the schedule in your book, do the job cold and torqued to spec when it comes due, and ignore anyone selling you a performance gain from a plug that has not worn out yet.

AAA puts spark plug replacement at $100 to $250, noting it can exceed $500 for larger engines or dealership work. RepairPal's estimator puts the same job at $284 to $397. AAA's low end and RepairPal's low end differ by 2.8 times for what is nominally the same service, and we are not going to split the difference — they are measuring different things. RepairPal's figures are derived from shop and dealer quotes and price OEM-equivalent long-life plugs at shop markup; AAA's are consumer guidance describing the straightforward case.

The dominant variable is not parts, it is access. Plugs themselves run roughly $3 to $25 each depending on type, and most engines need one per cylinder. Labour is where the money goes: an inline four with the plugs on top is 45 to 60 minutes, while a transverse V6 needs the upper intake manifold removed to reach the rear bank and runs two to three hours.

A Silverado costs roughly 2.2 times a Camry for the same nominal service, and nothing about the plugs explains that. It is entirely a function of how long it takes to reach them.

If a misfire has been left running long enough to damage the converter, the relevant number is a different one entirely — see catalytic converter replacement cost. That is the concrete price of ignoring a P0300, and it is the strongest financial argument for servicing on time.

This topic has more confidently-stated unsourced numbers than almost any other maintenance question, so it is worth being explicit about the gaps in this page rather than papering over them.

• No copper/platinum/iridium consensus table. It is the single most reproduced artefact on this subject and no plug manufacturer publishes it. We have given you what NGK and DENSO actually say instead, and those two do not agree with the table or with each other.
• No Bosch, Champion, Autolite or ruthenium figures. We could not retrieve first-party service-life claims for any of them. The per-material table circulating under Bosch's name traces to a reseller blog.
• No Ford, BMW, Mercedes-Benz or Honda mileage intervals. Ford's owner-manual server would not respond and its published figures conflict; BMW and Mercedes appear only in hedged dealer blogs and BMW may not publish a fixed figure at all; Honda uses condition-based Maintenance Minder rather than a mileage. The one Honda document we did retrieve was for a natural-gas Civic GX and cannot be generalised.
• No fuel economy or horsepower gain figures. Widely quoted, never sourced to a test.
• No single average cost. Published estimates differ by 2.8 times at the low end and real quotes reach three times the high end.
• No per-vehicle interval table. Of seventeen model-year combinations researched, only the 2021 Camry rows reached even medium confidence, and those came from a dealer quoting the schedule rather than from the manual.

One genuinely useful figure did survive from a Honda-published document, with a caveat attached. The owner's manual for a Civic GX specifies plug replacement at 100,000 miles, 160,000 km or 5 years under normal conditions, and 90,000 miles or 5 years under severe conditions. The mileage does not transfer to any other Honda, but the five-year cap is worth noticing: it is a time limit that almost no page on this subject mentions, and for a low-mileage commuter it binds long before the odometer does.

For where plugs sit in the wider service picture, see the car maintenance schedule, the 30/60/90k mile service where turbo and V6 plugs typically fall due, and the 100,000 mile service where long-life plugs usually land.