Mazda timing belt

My mother has a 2002 Mazda Millenia with 67K miles on it. I told her she should probably have the timing belt replaced, but her local Mazda dealership said she shouldn’t worry about it until the car hits 100,000 miles.

Having had a timing belt break while I was driving (it freakishly occurred to my 1999 Mazda Protege at 40K miles), I keep trying to convince her she doesn’t want a repeat of my experience. Should she listen to me or the dealership which, interestingly, is closing down at the end of next week?

Use The Change Interval That Is In The Owner’s Manual That Came With The Car.
There Is A Miles Recommendation And There Should Be An Age Recommendation.

Change the belt when the car reaches the miles or age, whichever occurs first. Since Mother’s Mazda is 8 to 9 years old, I’m quite certain it is past due for changing based on age. 2002 Mazdas were probably manufactured between 8/01 and 7/02.

Having a timing belt break at 40,000 miles is very unusual unless it was beyond the recommended age for replacement. Was your belt old or did warranty cover the premature belt failure ?

While car shopping for you next car you can look for cars manufactured without a timing belt. Cars with timing chains instead, that are well-maintained, often run their entire life without need for replacement.


Why are both of you ignoring the instructions provided by the manufacturer?

Yes, it was very unusual and Mazda said so, but they refused to cover it under the warranty. That was not a happy day. In the future I will certainly look for a car w/o a timing belt or one with a timing chain. Thanks for your speedy reply!

What could the manufacturer possibly know?
They are just the ones who designed, developed, and tested that engine prior to its introduction.

Surely a service writer (especially one at a dealership that is going out of business), and a car owner know much more than the manufacturer!


Just my opinion, but a timing belt should be changed at the 6 year mark no matter what the mileage is; and especially so if the engine is an interference fit one.

When customers speak with service people at the dealerships they’re generally speaking with a service writer. Very very few service writers have much mechanical aptitude or knowledge.

Your mother’s car was likely built in 2001 (verifiable by the tag on the drivers door jam) so this means the belt is approximately 9 years old. It’s way past due.

A belt popping at 40k miles is rare but it does happen with any make of car. Some good friends of mine suffered engine damage when their 3/4 year old Honda with 59k miles snapped a belt just a couple of weeks after buying the car. “AS IS” left them on the hook for this expense.
Some years back a 30k miles belt on my oldest son’s Ford snapped but at least in this case it was a non-interference engine and no damage was done.

Pay a couple hundred now, or a few thousand later

Both Engine Options In This Millenia May Be Free-Wheeling Designs.

If this is verifiable then additional engine damage probably won’t result from a breaking timing belt. However, the car will quit running immediately and unexpectedly and will not restart until the belt is replaced. This could put one in a dangerous situation.


The Gates timing belt guide lists 2 engines for the 2002 Millenia, a 2.3 liter engine( I think it is a sterling cycle) v-6 that is supposed to mbe changed at 60000 miles but is a non-interference engine and won’t be damaged if the belt breaks. The other engine is a 2.5 liter v-6 that is an interference engine and is supposed to be changed at 105000. I would guess that the 2.5 liter engine is the California engine because California requires the manufactured to pay for the belt change if it needs to be done in less than 105000 miles. I don’t think the California belts are better, just that the manufactures are willing to take their chances rather than pay for changing all the belts at 60000.

Oldtimer 11, Help Me Out Here. We’re Getting some conflicting Timing Belt Information.

The specific make / model / year / engine Gates “Look-up” to find out if the engine is of interference design says, “The vehicle shown is not an interference engine unless noted “Interference Engine Application” in the comments column of the Cam Belt application section.” . . .

. . . and in that column the only comment I see for both engines says, " Does not include cam. tensioner hyd. assy."

Also, in the downloadable guide book I see no Mazda Millenia models / engines / years listed as interference design. I’m thinking that both available engines for the 2002s are free-wheeling, but as I said if one was to take a chance on one breaking a timing belt then verification through Mazda would be advised.

What am I missing ?


I’m handicapped by actually knowing a little, not a lot, but more than most folks, about reliability measurement.

Some basics – One of the many ways you can subdivide engines is between those where the valves can not contact the pistons no matter what happens (non-interference) and those where they can (interference). A broken timing belt in a car with a non-interference engine is a nuisance and probably costs a few hundred more than a preventive replacement (towing, replacement rental, etc). A broken timing belt in an interference engine is a disaster – probably requiring an engine rebuild or replacement. You’ll find some posts in the thread about whether your engine is an interference engine or not. Probably not, but it’s somewhat unclear.

With regard to reliability. It is assumed that timing belts – like many other things follow what is called a “bathtub curve”. That is to say that there is a relatively high initial failure rate due to manufacturing defects, etc. Failures then fall away fairly quickly to a constant low failure rate. Finally, as the belts age and wear, the failure rate starts to creep up. If you plot number of failures vs time or miles, the curve is U-shaped. It looks vaguely like a bathtub.

The problem is that the right hand side of the curve is kind of hazy. The engineers won’t have that much data on timing belt failures from old age, and what data they have is for timing belts manufactured 6, 7, … 10 years ago – possibly with different materials or processes – and maybe for a somewhat different engine. Moreover, I’m told that the marketing folks (a dubiously necessary feature of modern society) want the longest practical replacement interval for timing belts.

So you’ll find a replacement interval specified for timing belts that is sort of negotiated between engineering and marketing neither of whom probably knows the “right” answer all that well.

My feeling. For interference engines, be meticulous about changing timing belts on or before the recommended date or miles – whichever comes first. For non-interference engines it’s probably OK to push the interval if you are tolerant of being stranded and have a good reason – e.g. you’re going to sell the vehicle soon or it has reached end of life, and you intend to scrap it the next time it breaks.

If your mother plans to keep the car for years and years, she’s probably going to have to replace the timing belt sooner or later anyway. Why not do it now?