A bike maintenance schedule you'll actually keep

Most maintenance advice fails for the same reason most training plans fail: it describes an ideal week that nobody has. You do not have two hours on a Sunday. You have ten minutes in a cold garage and a ride starting in twenty.

So this is a schedule built around what you will actually do, sorted by how often it needs doing and how long it takes. The five-minute jobs are at the top because they are the ones that matter most per minute spent, and the annual jobs are at the bottom with an honest note about which of them belong in a shop.

There is one piece of arithmetic in the middle that is worth more than all the rest of it put together, and it is about when you replace a chain.

The schedule

If you only ever do the first two rows, you will still be ahead of most riders. The one-minute check prevents the ride-ending failures. The weekly wipe-and-lube governs how fast everything in the drivetrain wears, and the published work on bicycle drive efficiency is consistent that lubrication condition and contamination are among the larger variables in how much power the drivetrain eats.

Chain wear: the only measurement that pays you back

Start with what wear actually is, because the usual name for it is wrong. A chain does not stretch. The side plates are not under enough tension to yield. What happens is that each pin and the bush it turns inside wear against each other, and the material removed adds to the distance between one pin and the next.

Wear modelling of roller chains, and measurements of pin and bush wear specifically, describe the same process: metal is removed at the pin–bush interface, the effective pitch grows, and the chain no longer matches the pitch of the sprocket teeth it sits on. Past a certain point the load stops being shared across several teeth and is concentrated on one or two, which is precisely how a worn chain machines a new profile into your cassette.

How to measure it

A steel rule beats most gauges. Chain pitch is half an inch, so twelve links measure exactly 12 inches when new. Lay a rule along the top run with the chain under light tension, line up the zero with the centre of one pin, and look at the pin 24 pins along:

• Exactly 12 inches — new.
• 12 and 1/16 inch — 0.5% elongation. Replace now.
• 12 and 1/8 inch — 1.0%. Expect to need a cassette too.

One warning about gauges. Many chain checkers work by seating a pin into a roller at each end and pushing them apart, which measures pitch plus the wear in the rollers themselves. Roller wear is real but it is not what damages your cassette, so these tools tend to condemn a chain earlier than a rule does. That is not a disaster — you are replacing a €40 part slightly early — but it is worth knowing which number you are looking at.

The cost argument, worked

Here is why 0.5% is the number. Assume €40 for a chain, €80 for a cassette and €110 for a set of chainrings, over 20,000 km.

The counterintuitive part is the middle of that list. Running chains longer means buying fewer chains, and it still costs more, because each one takes a cassette with it. The component you were trying not to buy is the cheapest thing in the system.

Two caveats so this is not overclaimed. The mileages are illustrative rather than measured — chain life varies by a factor of several between a waxed chain in summer and a wet-lubed one through a British winter, which is the subject of chain waxing vs lube. And the 0.5% threshold is a long-standing workshop convention rather than a figure from a published study; what the research supports is the mechanism, not the exact number.

Brakes

Pads and rotors

Disc pads are consumable and the limit is on the friction material, not the whole pad. Look through the caliper with a torch: when the material is down to roughly half a millimetre on either pad, replace them. Going further risks the backing plate contacting the rotor, which ruins a rotor to save a pad.

Rotors have a minimum thickness stamped on them, commonly 1.5 mm against a new thickness of 1.8 mm. A vernier caliper takes ten seconds and tells you whether the rotor or the pads are the reason the lever is coming further back.

Heat, which is what actually limits them

Dynamometer and field testing of bicycle disc brakes shows rotor and caliper temperatures climbing substantially during sustained braking on a descent, which is the real-world case that pad and fluid specifications are fighting. Two practical consequences: brake in firm intermittent applications rather than dragging, so the rotor gets time to shed heat, and treat a long alpine descent as a different duty cycle from anything you do at home.

Fluid

If your brakes use DOT fluid, it is hygroscopic — it absorbs water from the air, and absorbed water lowers the boiling point. This is well enough recognised that the US federal motor vehicle standard for brake fluids specifies both a dry and a wet equilibrium reflux boiling point, precisely because the fluid in service is not the fluid in the bottle. Annual replacement is a reasonable interval for a bike ridden in the wet.

Mineral oil systems are not hygroscopic, which is genuinely an advantage, but the oil still degrades with repeated heating and the system still admits air at seals. A spongy lever is the symptom either way.

Cables, housing and bearings

Mechanical shifting and braking degrade almost entirely through the housing rather than the cable. Outer housing fills with water and grit, the liner scores, and the friction rises until the return spring can no longer pull the mechanism back. The failure presents as sluggish upshifts and gets blamed on the derailleur. Replace the full run annually on a bike ridden through winter; it costs very little and restores shifting more reliably than anything else you can do.

Inner cables fail by fraying, almost always at the anchor bolt or inside the lever where you cannot see them. If a shifter feels notchy or a brake lever has gone gritty, look there before anything else.

Bearings have a grease life, not an infinite one. The tribology literature on grease-lubricated rolling bearings is clear that the grease, not the steel, is usually what determines how long the bearing lasts, and that contamination and temperature dominate that life. On a bicycle this translates into one very specific instruction: do not point a pressure washer at a hub, bottom bracket or headset. Water forced past a seal is the single most effective way to destroy a bearing, and it is entirely self-inflicted. Wet riding does the same thing more slowly, which is why bearing checks belong in the same session as everything else you do after a bad winter — see riding in the rain for the rest of that list.

What genuinely needs a shop

• Frame inspection after a significant crash. Bicycle frames are designed and certified against fatigue test protocols, and failures initiate at welds, joints and stress concentrations — places that are often under paint or inside a tube. Carbon damage in particular can be invisible from outside. A shop with the right experience, or a specialist inspection, is worth it before you ride a crashed frame hard again.
• Anything requiring a press or a facing tool. Headset cups, press-fit bottom brackets and disc mount facing all need tools that cost more than having it done twice a decade.
• Wheel building and serious truing. Getting a wheel round is easy; getting it round with even spoke tension is not, and uneven tension is why rebuilt wheels go out again.
• Suspension service. Seals, oil volumes and air spring dimensions are model-specific and unforgiving.
• Your first hydraulic bleed, unless you have the correct kit for your system and half an hour of patience. It is a genuinely learnable job, but doing it badly leaves you with worse brakes than you started with.

Two things that do not need a shop and are frequently handed over anyway: fitting a chain, and setting up a rear derailleur. Both are twenty-minute skills that pay for themselves in the first month.

The one-line version

Check tyres and brakes before every ride. Wipe and lube the chain weekly. Measure the chain monthly and replace it at 0.5%. Do a proper hour on it each season, and either do or buy the annual service. That is the whole schedule, and it is short enough to stick to — which matters far more than any individual item on it.

Maintenance is also the only performance work that never gets harder. A clean, correctly worn drivetrain, brakes that stop when you ask, and a power meter you have actually calibrated are all available on a Tuesday evening, for free, from a rider in any condition. Very little else in cycling can say that. And correct tyre pressure, which takes thirty seconds, is worth more watts than most of the components you have thought about upgrading this year.