Why the same printer strings in PETG and not in PLA
PETG is an amorphous copolyester. The glycol in the name is a modifier added to PET specifically to stop it crystallizing, which is what keeps it clear and easy to extrude, and it also changes how the melt behaves on the way out of the nozzle. PLA has a fairly sharp transition: it goes soft, flows, then firms up again over a narrow band. PETG stays soft and stretchy across a much wider one, with a glass transition somewhere around 80 to 85 °C against roughly 60 to 65 °C for PLA. Prusa’s own guidance puts the practical ceiling for PETG parts in the same place, rating them for outdoor use below 80 °C.
That difference is the whole story. When the nozzle lifts off a wall and travels, a small amount of material comes with it either way. On PLA the strand cools past its transition quickly and snaps. On PETG it is still in that soft, tacky state most of the way across the gap, so it draws out into a thread and welds itself onto the far side. You are not looking at a printer that has gone out of tune. You are looking at a material doing exactly what its chemistry says it will do.
Zero strings is the wrong target
Polymaker says this plainly in its own troubleshooting documentation, and I think it deserves more attention than it usually gets: some stringing on PETG is close to unavoidable, and the company tells you to plan on cleaning up the remainder with a razor blade or a heat gun rather than tuning it to nothing.
That reframes the job. You are trying to get from a print covered in visible webs to a print with a few fine hairs that come off with one pass of a heat gun. Chasing the last hair usually means dropping the nozzle temperature below where layers bond properly, or pushing retraction far enough to cause a clog, and both trades are worse than the strings were. If your general approach to stringing is not landing, the method in Reduce Stringing still applies, one variable at a time. What follows is the part that method leaves out: the settings where PETG behaves differently from everything else.
Moisture comes first, and it is not close
Every manufacturer that documents this puts damp filament at the top of the list. Bambu Lab’s wiki names it as the first cause of stringing and oozing. Polymaker lists drying or swapping the spool among its five fixes. Prusa’s advice is to simply try a different spool, on the grounds that the one in the machine has probably picked up water.
PETG absorbs moisture faster than PLA, which means a spool that sat open on a shelf through a humid summer can string badly at settings that worked fine in March. The tell is audible before it is visible: popping or crackling at the nozzle as trapped water flashes to steam. If you hear that, no retraction value is going to save the print. Wet Filament Symptoms covers how to confirm it against a known dry spool before you start changing settings you did not need to change.
Nozzle temperature, and where lowering it stops helping
Prusa’s stringing article gives a specific step: drop the nozzle by 5 to 10 °C and reprint the same test. Bambu’s wiki reaches the same conclusion from the material side, describing a nozzle running too hot as one that thins the melt until it flows when it should not.
The floor is layer bonding. Prusament PETG is specified at 230 °C for the first layer and 240 °C after, and PrusaSlicer’s own newer profile for that same filament goes as high as 250 °C, which tells you how wide the usable band is even within one brand. Go far under it and you will trade visible strings for layers you can pull apart with your fingers. Test the piece, do not just look at it.
Nobody agrees on a retraction number, and that is the honest answer
This is where most PETG advice falls apart, because the published figures genuinely conflict and none of them is wrong.
Prusa caps retraction length at 2 mm on the MK2.5 and MK3 family, and ships the Bowden MINI with a default of 3.2 mm, because the several centimeters of tube between gear and nozzle compress under pressure and have to be accounted for. Bambu Lab recommends staying at or under 2 mm on its direct drive machines and says so specifically to avoid plugging the hotend. OrcaSlicer’s wiki notes that PETG and TPU are more prone to stringing and may want longer retraction than PLA or ABS, and then declines to publish a number at all.
Those three positions are consistent once you notice they are describing different hardware. Retraction length is a property of the gap between the drive gear and the melt zone, not of the filament. What the filament changes is how much of a penalty you pay for getting it wrong. The practical move is to run a retraction test on your own machine with the spool you actually have loaded, which OrcaSlicer includes as a built-in calibration, rather than copying a value off a forum post from someone with a different extruder.
Retraction speed has the same shape of answer. Prusa’s guidance is that raising it helps stringing until the extruder motor starts skipping steps, at which point you have gone too far. There is no number that survives being moved to another printer.
Z-hop usually makes PETG worse, and turning it off has a cost
This one surprises people. Lifting the nozzle before a travel move sounds like it should help, and on PETG it often does the opposite. Prusa states it directly in its stringing article: a lower lift Z setting improves stringing. The reason is mechanical. The nozzle is already oozing when the lift happens, so raising it stretches the strand into the air and gives it clean room to draw out, instead of dragging it through material where it would have been buried.
Prusa also attaches the warning that belongs with this advice, and it is worth taking seriously: with Z-hop disabled the nozzle can collide with the printed part. That failure has its own signature, and if you turn Z-hop off and start seeing gouges or scrape marks across the top of your prints, Fix Nozzle Dragging covers what you are actually looking at. On a tall model with thin features, keeping a small Z-hop and accepting a few more strings is often the better trade.
Travel routing, plate layout and speed
Strings form during travel, so anything that shortens travel or hides it inside the part helps. Both PrusaSlicer and Bambu Studio expose the same two ideas under different names. Prusa has Avoid crossing perimeters under Layers and perimeters, plus Only retract when crossing perimeters under Infill, which lets any ooze land inside the walls where nobody sees it. Bambu’s equivalent is Avoid crossing wall, and its wiki lists enabling it as a direct way to cut stringing.
Bambu adds a point that rarely shows up in stringing advice and matters more on PETG than on anything else: how you arrange parts on the plate. Many small separated objects, or a few objects spread far apart, mean long travel moves with a soft melt hanging at the nozzle the whole way. Bambu’s own recommendation is to reduce the spacing between models. It costs nothing and it removes the travel moves rather than trying to survive them.
Faster travel helps for the same reason, since less time in the air means less thread. Polymaker suggests 200 mm/s and above with travel acceleration around 2,500 mm/s² on a CoreXY frame, and points out that you are not printing during those moves, so there is no surface quality to lose. Older bedslinger frames will rattle or skip well before those numbers. Back off when you hear it.
Coasting and wipe, which PETG rewards more than PLA
Coasting replaces the tail end of an extrusion path with a travel move, letting pressure already in the nozzle finish the line instead of adding more plastic. Polymaker recommends trying it and says the gain is most noticeable on PETG and on Bowden machines. The side effect is real though: leave coasting on when you did not have a stringing problem and you get small holes in the sides of parts where the extrusion stopped short.
Wipe is the quieter one. Both PrusaSlicer and OrcaSlicer move the nozzle along the last printed path while retracting, which scrapes off material that would otherwise leave with it. Prusa recommends leaving it on. OrcaSlicer has since added a setting that splits the retraction across the wipe, some before and some after, which its wiki describes as helpful for detailed models and for filaments prone to stringing. That is PETG, if it is anything.
The film on the nozzle that no setting fixes
Prusa documents one cause that is specific enough to name PET-G directly, and it is the one people miss. Print long enough from a single material and it builds a thin layer on the nozzle. Strands then stick to that layer and get dragged onto the print, so the machine produces stringing that no slicer setting touches, because the source is a coating rather than the extrusion. Prusa’s fix is unglamorous: clean the nozzle thoroughly before printing, and clear out remnants of whatever ran before.
If you have already fixed moisture, dropped the temperature, tuned retraction and are still getting hairs, look at the outside of the nozzle before you change anything else.
The cooling trade-off you cannot settle with a setting
Here is the contradiction, and I would rather name it than pretend it resolves. Prusa says cooling the PETG print helps keep detail and prevents stringing and oozing. Prusa also says that if you want the part as tough as possible, turn the fan off, because more heat merges layers better. Its middle-ground recommendation is fan off for the first few layers, then about half power.
Both statements are true, and they pull opposite ways. More cooling firms the strand sooner and cuts stringing. Less cooling gives you the layer adhesion PETG is usually chosen for in the first place. There is no fan speed that maximizes both, so decide what the part is for before you decide the number. A display piece can take full cooling and a few weak layers. A bracket that has to hold something cannot. PETG Print Settings has the full starting profile if you want the other numbers alongside this one.
Most PETG stringing complaints resolve into damp filament plus a temperature 10 °C higher than it needed to be. The rest is trade-offs, and knowing which trade you are making beats finding a magic value that only ever worked on somebody else’s printer.