The short version
- Bearings have no replacement interval. They are replaced on condition, never on mileage.
- In one shop survey, 49 percent were found by someone inspecting, not by the driver noticing.
- Nearly a quarter turned up during a brake job, which is the cheapest place to find one.
- Doing a bearing at the same time as brakes saves roughly half an hour of labor per corner.
Knowing when to replace a wheel bearing differs from most maintenance decisions, because there is no schedule to follow.
Brake pads wear predictably. Oil degrades on a timetable. A bearing does neither, so the decision is about what you find rather than what the odometer says.
When to replace a wheel bearing: never on mileage alone
Worth being firm about, because people want a number. There is no mileage at which to replace a wheel bearing.
A sealed hub averages 85,000 to 100,000 miles, but that average hides a huge spread. Plenty fail at 40,000 and plenty are original at 200,000.
Replacing a healthy bearing costs full price and buys nothing. The new one starts its own uncertain life from zero.
Worse, fitting is where most premature failures come from. A preventive swap adds installation risk while removing no known problem.
The exception is when the labor is already being spent, which is the whole point of the section below.
How bad bearings actually get found
Why no service schedule lists them is on the inspection page.
A shop survey of how failures came to light produces a useful breakdown.
| How it was found | Share |
|---|---|
| Customer reported a noise | 51 percent |
| Found during a brake job | 24 percent |
| Found during an alignment | 19 percent |
| Other | 6 percent |
Read the bottom three rows together. Roughly 49 percent were found by somebody looking, before the driver noticed anything.
That is the argument for inspection. Half of all failing bearings are catchable during work you were already paying for.
The thresholds that mean replace
Five findings, and any one of them is decisive.
- Roughness or grinding when spun by hand. The raceways are already damaged and it does not improve.
- Measurable play beyond specification. On a dial indicator, more than a few thousandths on a sealed hub.
- Noise that changes with load transfer. A hum that grows on one curve and fades on the other is close to conclusive.
- A hub running noticeably hotter than its opposite number. More than about 20 degrees Celsius between the same points.
- Grease thrown onto the inside of the rim on a serviceable set, meaning the seal has failed.
How to perform each check is on the diagnosis page. The play figures are on the play and vibration page.
The findings that mean monitor
Equally important, because unnecessary replacement is a real cost.
A faint hum with no other symptom. Confirm the bearing first, since cupped tires sound nearly identical and are far more common. The impostors page covers the rest.
Barely perceptible play on a serviceable tapered set. That may be correct adjustment rather than wear, since those run with a thousandth or two of end play.
Slight warmth after a drive. Hubs get warm. The test is the comparison between sides, not the absolute figure.
High mileage with no symptoms. A quiet bearing at 150,000 miles is a working bearing.
The one case for replacing early
This is where the labor overlap changes the arithmetic.
A brake job and a front hub replacement share most of their steps. Wheel off, caliper off, bracket off, rotor off, all done once either way.
| Scenario | Labor |
|---|---|
| Brake job alone | About 0.5 to 0.7 hours |
| Bolt-on hub alone | About 1.0 to 1.5 hours |
| Both, done separately | About 1.5 to 2.2 hours |
| Both, done together | About 1.1 to 1.6 hours |
At common shop rates that is around $50 to $85 per corner, purely from not doing the same teardown twice.
So when to replace a wheel bearing shifts if the corner is already apart. A borderline one is worth doing now rather than in six months.
Borderline is doing real work in that sentence. Act on a genuine finding while the corner is apart, but do not replace healthy parts on principle.
When one side fails
The question that follows immediately, and the answer is usually no.
Bearings are not a matched pair the way brake pads are. The two corners share almost no labor, so doing both costs close to double.
Two situations change that. The other side is already making noise, or the car is high mileage in a salt state where both corners have had identical lives.
Even then, check the second side rather than assuming. The lifespan page covers why the passenger side usually goes first.
Waiting, and what it costs
A noisy bearing is not an emergency, but waiting is not free either.
Once material comes off the raceways, every mile circulates debris and accelerates the damage. A bearing that is merely humming can end up damaging the knuckle bore.
That escalation is what matters financially. A bearing is a few hundred dollars, and a bearing plus a knuckle is considerably more. The bearing or whole hub page prices it.
How long you actually have, and what changes the urgency, is on the driving page.
The financial case for acting early is on the collateral damage page, since almost everything expensive happens late.
Common questions
So when should I replace a wheel bearing?
On condition, never on mileage. Roughness when spun, play beyond specification, noise that changes with load transfer, or a hub running hot compared to its opposite number.
Is there a replacement interval?
No. Sealed hubs average 85,000 to 100,000 miles but the spread is enormous, so mileage alone never tells you when to replace a wheel bearing.
Should I replace one preventively?
Generally no. It costs full price, buys no certainty, and introduces the installation risk that causes most early failures.
What if I am already having the brakes done?
Then a borderline bearing is worth doing now. Deciding when to replace a wheel bearing changes once the teardown is already paid for, and the overlap saves $50 to $85 per corner.
Do I replace both sides together?
Usually not. They are not a matched pair and share almost no labor, so doing both costs close to double rather than a little more.
How are most bad bearings found?
About half by the driver reporting a noise. The rest turn up during brake work and alignments, which is why inspection during other jobs is worth asking for.
What happens if I wait?
The damage accelerates, because debris circulating in the bearing keeps working on the raceways. The expensive escalation is damage to the knuckle bore.