Retraction is the slicer move that pulls filament back slightly before the nozzle travels across an open gap. The goal is simple: reduce ooze and strings without retracting so far or so aggressively that the next line starts weak, late, or inconsistent. The right value is specific to the printer, hotend, filament, nozzle temperature, and filament path—not a number to copy blindly.
This guide gives you a controlled way to establish a starting point, run a useful test, and identify when retraction is not the problem. It applies to Cura, OrcaSlicer, PrusaSlicer, and other slicers with comparable distance and speed controls.
🔴 Important
Do not use retraction to hide wet filament, a partially clogged nozzle, or a temperature that is too high. Those conditions can create the same wisps and blobs, but a larger retraction value will not solve their root cause.
What retraction settings actually control
| Setting | What it changes | Why it matters |
|---|---|---|
| Retraction distance | How far filament is pulled back before a travel move | Too little can leave wisps; too much can cause delayed restart, grinding, or inconsistent extrusion. |
| Retraction speed | How quickly the extruder pulls filament back and pushes it forward | Too slow may not control ooze; too fast can stress the filament path or create inconsistent behavior. |
| Minimum travel / travel threshold | When the slicer is allowed to retract | Prevents needless retractions during tiny moves that add print time and wear. |
| Z hop | Whether the nozzle lifts while travelling | Can protect a surface or avoid a collision, but is separate from the amount of retraction. |
The core settings to change deliberately
Direct-drive and Bowden systems need different starting points
A direct-drive extruder places the drive gears close to the hotend, so there is less flexible filament path between the gears and melt zone. A Bowden setup has a longer tube between those points. That difference is why a profile that behaves well on one design can be wrong for the other. Start from the profile supplied for your exact printer when it is available, then tune with a test rather than borrowing a number from a different machine.

If you need a quick baseline before testing, use the Retraction Settings Guide. It is designed to give a sensible start, not a replacement for a controlled test on your own printer.
Set the baseline before you tune
Retraction is easier to diagnose when the other high-impact variables are stable. Load the material profile, confirm the filament diameter, keep the nozzle temperature in the material’s normal range, and check that the filament is feeding without a snag. If the spool has been exposed to humid air or shows popping, roughness, or dramatic new stringing, dry it first and compare one change at a time.
Use the Filament Drying Guide and Filament Temperature Chart to establish those material baselines before chasing retraction.
A simple retraction calibration workflow
1. Use one familiar filament and one stable profile
Choose a spool that is feeding cleanly and a profile you have already used successfully. Do not compare a new wet PETG spool, a new nozzle, and a new retraction value in the same test. The goal is to make the next result easy to interpret.
2. Print a test with repeated travel moves
A retraction tower or paired-pillar test is useful because it creates open travel moves repeatedly. Look for wisps between pillars, blobs at restart points, missing material just after travel, and any clicking or grinding from the extruder. A single decorative model with lots of changing geometry is harder to compare.

3. Change one setting in small steps
Start with distance or speed—not both. Run the same test after one small adjustment, note the change, and keep the better result. Once you find a clean range, make one final confirmation print rather than pushing past the point where strings disappear. Over-tuning can introduce weak restarts or more wear without improving the part.
4. Verify on the part you actually want to print
A tower only shows travel behavior. Print a small version of the real object next, especially if it has many seams, thin features, or long travel moves. If the final part develops a different issue, work from the symptom rather than immediately increasing retraction again.
How to read the result
| What you see | Likely direction | Check before changing more settings |
|---|---|---|
| Fine wisps between open features | Retraction may be too low, or temperature/moisture may be contributing | Dryness, nozzle temperature, travel behavior |
| Gaps or weak lines after travel | Retraction may be too high or restart behavior too aggressive | Extruder grip, nozzle condition, flow profile |
| Blobs at restart points | Retraction, wipe/coast equivalents, temperature, or pressure behavior may need attention | Use the same test and isolate one setting |
| Clicking or chewed filament | Do not keep increasing retraction | Partial clog, heat creep, spool drag, extruder tension |
Common retraction-test outcomes
When stringing is not a retraction problem
Stringing is a visible symptom, not a setting category. Wet filament, excess nozzle temperature, a slow or poorly planned travel path, and a poorly tuned filament profile can all contribute. Make retraction the next controlled step only after the material and extrusion path make sense.
Use the full 3D print stringing troubleshooting guide for the right diagnostic order. For PETG specifically, the PETG stringing guide explains why material condition and temperature are often part of the answer.
Retraction settings FAQ
Should I use the same retraction settings for every filament?
No. Even on the same printer, materials can behave differently at their normal temperatures. Keep a tested baseline per material and refine it only when a controlled test shows a repeatable need.
Does increasing retraction always reduce stringing?
No. Too much retraction can produce weak restarts, gaps, filament grinding, or heat-creep-related problems. Stop when the result is clean and reliable rather than aiming for the largest possible value.
Should I tune temperature or retraction first?
Set a reasonable material temperature first, confirm the spool is dry and the nozzle is healthy, then tune retraction. This sequence gives the retraction test a clearer signal.
Next step
Use the tool for a starting point, record the result for your printer and filament, and keep the successful profile with the spool or slicer preset. That turns a one-time cleanup into a repeatable workflow instead of a stringing emergency before every print.
Print quality workflow
Calibration & slicer settings
Build a reliable baseline before chasing individual settings: start with the printer, then tune material, flow, and slicer quality settings one controlled step at a time.
- 01Cura ProfilesStart with a sensible profile baseline before making material- or printer-specific adjustments.
- 02Z-Offset Wizard: Complete GuideSet nozzle height correctly so the first layer has the right amount of squish.
- 03Cura First Layer SettingsTune the settings that turn a good Z offset into a consistent first layer.
- 043D Printer E-Step Calibration: A Practical BaselineEstablish a supported hardware feed baseline before tuning a material-specific flow setting.
- 053D Printer Flow Rate Calibration: Measure Before You AdjustUse a repeatable wall measurement to make small, material-specific flow corrections.
- 06Cura Temperature Tower: Find a Better Starting TemperatureCompare one spool across a controlled temperature range before changing other slicer settings.
- 073D Printer Retraction Settings: Distance, Speed, and CalibrationYou are here
- 08Cura Speed Settings: Balance Speed, Quality, and ReliabilitySeparate outer walls, infill, travel, first layers, and acceleration to find a reliable setting.
- 09Cura Jerk SettingsRefine motion behavior after the basics are stable and repeatable.
- 10How to Hide Z-Seam Imperfections in CuraUse finish-focused adjustments after calibration and material settings are working together.
Fix it systematically
Choose the next useful check
Work from the visible symptom to a controlled test instead of changing temperature, flow, and mechanics at the same time.
