I have a 1996 Jeep Cherokee Sport with 153K miles. I’ve been on a maintenance kick lately and I’m wondering, is it reasonable to start replacing sensors before they go bad? I’ve replaced both oxygen sensors over the last few years but what about the others, MAPS, TPS, crankshaft position sensor, etc? The Jeep is running great but I’ve been driving a lot more for work and just want to keep it that way. Thanks!
I’ve NEVER EVER seen anyone replace an O2 sensor as a PM item. I’ve had vehicles go well beyond 300k miles on the original 02 sensors…Same with other sensors…replace when needed. The ECU will tell you when there’s a problem.
Nope!
Neither have I.
Depending on the vehicle, these sensors can get pretty expensive.
Let the On Board Diagnostic system inform you if there’s a problem with a sensor.
After all, that’s what it’s designed to do.
Tester
Plus, there is the chance that the replacement sensor could be of poor quality and die. Leave well enough alone.
I’m Relatively Certain That If You Began Replacing Parts And Replaced Every Part On The Car Except One, According To The Laws Of Murphy, That’s The Part That Would Break.
CSA :neutral:
Thanks all! I should clarify,I had a CEL which prompted me to change the oxygen sensors. I appreciate the input.
I don’t entirely agree about the mil telling you when an oxygen sensor needs to be replaced
I’ve seen a few lazy and/or skewed oxygen sensors which were not setting a code
Not even a pending code
I wouldn’t be surprised if a few of the other pros agree with me on this
OBD2 is pretty useful, but you can’t rely on it for everything
I’m not advocating that OP starts replacing sensors just because they’re old
What I am saying is that the absence of a code or mil does not mean a particular sensor is okay
And I’m applying that last statement to sensors in general, not just oxygen sensors
Always wondered about this myself, can an ohm test on an o2 senspr be used as an indication of it’s viability?
An O2 sensor is more like a battery than a resistor. There’s demos on U2 how to test O2 sensors to some degree using a propane torch and a voltage meter. Aslo be aware there’s different kind of O2 sensors, using different technology, so you need to know which kind you want to test.
I agree with db4690 as usual.
For what it’s worth, when O2 sensors first came into use some car makers did consider them a normal replacement maintenance item. SAAB said every 15k miles. Ouch…
O2 sensors back in the day were 400 bucks each and customers balked.
SAAB then said every 30k miles…
Customers balked.
SAAB then said every 60k miles.
Customers balked.
SAAB then said “as needed”.
Customers still balked. 
I’ve been wrong many times, and will be again . . . but thanks
Anyways, I suppose those $400 sensors were in the late 70s/early 80s . . . ?
I wonder how that would “translate” into 2016 dollars ?
I remember even into the early 1990s, there were a few Benzes where the “replace oxygen sensor” light would come on at set intervals
There was a procedure to reset some kind of module/relay, presumably after you replaced the sensor. But if I recall correctly, there were only so many times you could reset the thing, before you needed to install a new relay
And I also recall there was an even earlier model vehicle, where the factory procedure was to REMOVE the oxygen sensor light from the cluster 
It is not difficult to monitor oxygen sensor performance, Chrysler computers will trigger oxygen sensor slow response faults while other manufactures do not.
Since the 1980’s vehicles that could not pass the EPA 100,000 mile emissions durability test were required to illuminate a maintenance light to replace items that were known to fail. When vehicles came in with the emissions maintenance light on at 60,000 or 87,500 miles etc. for a maintenance light reset I would prepare an estimate for the necessary maintenance. The service writers would tell me to just reset the monitor as this pays .5 hours labor. They were unfamiliar with the emissions maintenance schedule because most of the mechanics that I worked with were illiterate and never looked up the maintenance necessary for the vehicle.
"most of the mechanics that I worked with were illiterate"
THAT is frightening…
Most the sensors the OP listed would not disable the vehicle except the crankshaft position sensor. Most would throw a CEL and keep right on running. Some would default to limp-home mode and some like a crankshaft position sensor would stop the car. If the “disable the car” sensors are easy to reach and replace, I’d be tempted to buy that sensor and keep it in my in-car toolbox so I could run it 'til it fails. Murphy’s law then says I’d never have to replace it.
If its a real bear to replace like some crank and cam position sensors, I’d consider replacing them early but only with sensors from the dealer, no Chinese knockoffs.
This is the conundrum all owners of older, high mileage cars have to face.
Denso and NGK both make high quality sensors…and are the OEM manufacturer for many/most/all of the Japanese companies.
I’d leave them alone. Along with the risk of replacing a working sensor with a bad one, or one that fails early, there’s the real risk of breaking something (plug, wire, etc.) when swapping out the sensor. No need to take that risk.
This is one case where the old adage makes sense: “If it ain’t broke, don’t fix it!”.
Nope, let the sensor do its job and then fail . . .then replace it. Rocketman
Probably over 1/2 of the ckp sensors I’ve replaced in the past few years have had absolutely no codes. Not even pending. But they tested bad, were causing the no start/stalling problems. And the problems were 100% resolved by replacing them
I’ve replaced several maf sensors that weren’t setting a code. Not even pending. Some caused the typical bogus lean/rich codes. Others caused a no-start condition
Just because there’s no mil and no codes, doesn’t mean it isn’t broke
I would argue that if there’s no code AND the engine’s runnin’ great, it ain’t broke.
I support your post, no code doesn’t necessarily mean it isn’t the cause of an operating problem, but the original question was about preemptive replacement.