PETG Stringing: The Fixes Ranked by What Actually Works

Your PETG part comes off the plate wearing a spider web. You do what everyone does: crank retraction distance, print a retraction tower, crank it more. The webs remain. By Thursday you’ve got 6mm of retraction on a direct-drive extruder, a heat-creep clog, and the strings are still there.

Here’s the thing the retraction-tower ritual misses: for PETG, retraction is fix number three. The two fixes above it matter more, and one of them can make stringing unfixable by any slicer setting while it’s active.

This is the ranked list — what to do, in order, with the actual numbers from Prusa’s and Bambu’s official docs, Ellis’s tuning guide, and the community’s collective scar tissue. Stop when the strings stop.

Fix 1: Dry the filament (yes, first)

PETG is hygroscopic, and wet PETG strings in a way no setting can compensate — the water flashes to steam in the hotend and pushes molten plastic out of the nozzle during travels. Ellis’s guide says it flat out: wet filament causes near-unfixable stringing; Bambu’s wiki lists “stringing and bubbles” as the first symptom of undried PETG.

The drying specs, per the people who make the stuff:

Source	Temp	Time
Prusa knowledge base	55°C	6h
Bambu Lab wiki	55°C (65°C for PETG-HF / PETG-CF)	8h
eSUN spec sheet	65°C	~3h

The spread (55 vs 65°C) is a real disagreement — 55°C is the safe universal number; 65°C dries faster but will deform cardboard spools and is only for filament you can re-spool or that ships on plastic. If your prints also show tiny pops, micro-bubbles, or matte patches, moisture is confirmed; dry before touching anything else.

Tell-tale that this was your problem: the same profile stringing worse in monsoon season than in January. PETG left out in 60%+ humidity picks up enough water to string within days.

Fix 2: Drop the nozzle temperature 5–10°C

Hotter PETG is runnier PETG, and ooze scales with runniness. Both Prusa’s stringing article and Simplify3D’s troubleshooting guide give the same first move: −5 to −10°C.

PETG’s usable window is roughly 220–250°C (Prusa’s own profile runs 230°C first layer / 240°C after). If you’re at 245°C, try 235°C. The clean way to find your filament’s floor: a temperature tower from 250 down to 220 in 5°C steps — pick the coolest band where bridges still look good, because every 10°C you remove also costs some layer adhesion. PETG parts that need strength shouldn’t print at the very bottom of the window; find the compromise band.

Fix 3: Retraction — small, slow, and less than you think

Now the setting everyone starts with. The numbers that work, by extruder type:

Extruder	Distance	Speed
Direct drive	0.5–1mm (start 0.5, tune in 0.1 steps)	30–35mm/s
Direct drive (Bambu stock reference)	0.8mm default; up to ~1.5 for PETG	30–40mm/s
Bowden	2–5mm (Ender 3 classic: ~6mm)	25–40mm/s

Two counter-intuitive findings from Ellis’s testing: slower retraction beats faster for PETG (35mm/s outperformed higher speeds — the melt needs time to relax), and beyond ~1mm on direct drive you’re not removing strings anymore, you’re pulling molten PETG up into the cold zone and manufacturing your next clog. If 1mm direct / 5mm bowden hasn’t fixed it, more distance won’t — the problem is upstream in fixes 1 and 2.

The old “up to 15mm for bowden” advice still floating around dates from pre-pressure-advance firmware; modern Klipper/Marlin with PA tuned needs a fraction of that.

Fix 4: Travel speed up, crossings off

Strings form during travel moves — so spend less time traveling. Raise travel speed (150 isn’t a typo on modern machines; 3DSourced suggests starting ~90mm/s and stepping up 10 at a time on older ones) and enable “avoid crossing perimeters” (PrusaSlicer) / “avoid crossing walls” (Bambu Studio): a nozzle that routes around the part instead of over it can’t decorate the part on the way.

Fix 5: Wipe and the small hygiene settings

• Wipe while retracting: ON — scrubs the nozzle tip on the wall it’s leaving instead of carrying a droplet to the next one. Prusa enables it in their PETG profiles.
• Retract on layer change: ON.
• Leave minimum travel for retraction at its ~1mm default — raising it to “save time” un-retracts exactly the short hops that string.

Fix 6: Z-hop — the setting the guides fight about

Genuine disagreement, worth knowing both sides: Ellis runs z-hop at 0.2mm and reports stringing above 0.3mm; much of the Bambu community turns z-hop off for PETG entirely, reasoning that every hop stretches a thread on tacky material; Bambu’s stock profiles ship with it on. The physics supports both — hopping avoids dragging the nozzle through ooze, but the vertical pull spins strings from soft PETG.

Practical resolution: if strings persist at this point in the list, toggle whatever your z-hop currently is and print the same test. It’s a two-minute experiment and the answer is specific to your machine.

Fix 7: Cooling — quality lever, weak stringing lever

PETG likes less fan than PLA (Prusa runs ~50%, Bambu’s PETG-CF spec says 0–40%). Some guides report dropping fan from 100% to 40–60% visibly reduces stringing on Bambu machines; others see no stringing effect at all. Consensus is only that 100% fan is wrong for PETG generally — set 40–60% for part quality and bridging, and treat any stringing improvement as a bonus.

The one-page checklist

In order, stopping when clean:

1. Dry: 55°C / 6–8h (65°C for HF/CF variants, plastic spools only)
2. Temp: −5 to −10°C, or run a 250→220 tower and pick the coolest clean band
3. Retraction: DD 0.5–1mm @ 30–35mm/s · bowden 2–5mm — tune in 0.1mm steps, never past the clog line
4. Travel: speed up, avoid crossing walls
5. Wipe on, retract on layer change on, min-travel at default
6. Z-hop: flip yours and compare
7. Fan: 40–60%

And once the part is clean but wearing three faint whiskers: 2 seconds with a heat gun (or a lighter, carefully) vaporizes hair-thin PETG strings without marking the part. Every production PETG farm finishes parts this way — sometimes the last fix is post-processing, not another slice.

Stringing sorted and warping next? PETG rarely warps, but if you’re also fighting first-layer problems, the first-layer diagnostic and PLA vs PETG guide cover the rest of the material’s personality.