Prusa MK4 & MK3S+ Maintenance Guide

Prusa publishes one of the most detailed official maintenance schedules in the hobby — this is the shortlist of what actually matters most.

Maintenance schedule

TaskInterval
Lubricate smooth rods (X/Y/Z)Every ~3 months or 500 hours of printing
Check belt tensionEvery ~3 months
Check extruder idler tensionEvery ~1–2 months
Clean/replace nozzleEvery 6–12 months, sooner with abrasive filament
Re-run first-layer / Live Z calibrationAfter any nozzle swap or noticeable first-layer change
Check probe (PINDA/SuperPINDA on MK3S+) for driftIf first layer height varies with ambient temperature

Why lubrication is the single most-cited maintenance item

Dry or under-lubricated smooth rods on the X/Y/Z axes increase friction, which shows up as inconsistent layer lines, ringing, or unusual stepper motor noise long before it causes an outright failure — Prusa's own manual calls out a schedule specifically because this is the maintenance item most likely to get skipped since the printer still runs fine for a long while as it degrades. A light, printer-specific grease applied per the official manual matters more here than on printers with sealed linear rails.

MK4's automatic calibration vs. MK3S+'s more manual process

The MK4 added an accelerometer for input shaper (resonance) calibration and generally automates more of the tuning process than the MK3S+, which relies more on manual Live Z adjustment and the PINDA/SuperPINDA probe for first-layer height. If you're used to one and moving to the other, expect the calibration workflow itself to be meaningfully different, not just the printer's speed.

Probe drift is a well-known, well-documented quirk

The PINDA and SuperPINDA inductive probes used on MK3S+ can shift their trigger point slightly with ambient temperature, which shows up as a first layer that's fine in the morning and slightly off by afternoon in a room with real temperature swings — this is a widely known characteristic of the sensor type, not a defect, and Prusa's firmware includes temperature compensation for exactly this reason. If first-layer height seems to drift with room temperature rather than anything you changed, this is the likely explanation.

Related Guides