The nozzle drags across your bed leaving a transparent smear. Or it hovers 0.3mm above the surface, laying down a round, unbonded filament tube that peels off on the second pass. The first layer is where 90% of print failures begin, and getting it right is about measurement, not guesswork.
The Target: What A Perfect First Layer Looks Like
A properly squished first layer has these characteristics:
– Flat on top (the nozzle’s flat face irons it smooth)
– Slightly wider than the nozzle diameter (0.45-0.48mm for a 0.4mm nozzle)
– No gaps between adjacent lines (they should merge at the edges)
– Consistent across the entire bed (no thin spots in the center, no thick edges)
Grab a calibration print — a single-layer 75×75mm square — and measure the thickness with calipers. If your first layer height is set to 0.2mm in the slicer, the printed square should measure 0.20mm ±0.02mm. Thinner means the nozzle is too close (over-squished). Thicker means too far.
Step-by-Step First Layer Calibration
1. Manual Bed Leveling (Tramming)
Before any Z-offset adjustment, the bed must be physically parallel to the nozzle’s X/Y motion plane. This is tramming, not “leveling” — the bed doesn’t need to be level to the ground, it needs to be parallel to the gantry.
Method (Marlin with manual leveling):
– Home the printer
– Disable steppers (M84 or use the menu) so you can move the head by hand
– Slide a piece of copy paper (0.10mm typical thickness) between the nozzle and bed at each corner
– Adjust the bed screw until the paper slides with slight resistance — enough to feel friction but not enough to tear
– Repeat all four corners in sequence (adjusting one corner tilts the bed, affecting the others)
– Go around at least twice, until all four corners feel identical
With a bed probe (BLTouch/CR Touch): The probe handles mesh compensation, but the bed still benefits from rough mechanical tramming. Use the probe-assisted tramming wizard if your firmware supports it, or tram manually first.
2. Set Z-Offset with Live Adjust
Start a first-layer calibration print (a large single-layer square). Watch the first lines go down. Adjust the Z-offset in real-time via your printer’s LCD or terminal:
Marlin: Use the “Babystep Z” or “Z Offset” menu during the print. Adjust in 0.02mm increments.
Klipper: Use the SET_GCODE_OFFSET Z_ADJUST=-0.02 command in the console during the print, or the Mainsail/Fluidd interface slider.
What to look for during the live test:
– Too high: Individual lines have rounded tops, don’t merge with neighbors, and lift off the bed with a fingernail. The square measures thicker than 0.20mm. → Lower Z by 0.05mm.
– Too low: Lines are transparent, have ridges (the nozzle plows through previously laid filament), and the surface feels rough like sandpaper. The square measures thinner than 0.18mm. → Raise Z by 0.05mm.
– Just right: Lines merge seamlessly at the edges, the surface is smooth, and the square measures 0.20mm.
3. Confirm with a Full-Bed Test
Print a single-layer pattern that covers the full bed surface (like the popular 5-spot or 9-spot test). Check all corners and the center. If the center is perfect but the front-left corner is rough (over-squished), the bed mesh isn’t compensating correctly. Re-run your mesh leveling and verify the probe is reading consistently.
4. Save and Verify
Save the Z-offset to EEPROM (M500 in Marlin, SAVE_CONFIG in Klipper). Restart the printer and run the test again. The offset should survive a power cycle. If it shifts after a reboot, your Z endstop or probe is inconsistent — check for loose mounting, debris on the probe pin, or binding on the Z-axis lead screw.
First Layer Diagnosis Table
| Symptom | Likely Cause | Fix |
|---|---|---|
| No adhesion, filament curls | Nozzle too far from bed | Lower Z-offset by 0.05-0.10mm |
| Transparent/smeared lines | Nozzle too close | Raise Z-offset by 0.05mm |
| Good corners, rough center | Warped bed (dish shape) | Enable bed mesh leveling |
| Good center, rough edges | Bed not trammed | Re-tram corners manually |
| Elephant’s foot (bulging bottom) | Nozzle too close + bed too hot | Raise Z 0.02mm, lower bed 5°C |
| Good adhesion on one side only | Uneven bed screws | Re-tram — one side higher |
| Fine first layer, later layers shift | Z binding or loose coupler | Check lead screw, coupler grub screws |
What Most Pilots Get Wrong About First Layer
Mistake 1: Using a brim or raft to compensate for bad first layer calibration.
A brim masks a bad Z-offset — it doesn’t fix it. If you need a brim on every print, your first layer is wrong. Fix the offset and you’ll print brim-free on 90% of models. Brims should be for models with tiny bed contact, not for compensating poor adhesion.
Mistake 2: Leveling with cold nozzle.
A hot nozzle has a tiny thermal expansion (0.02-0.05mm longer). If you tram with a cold nozzle then print at 210°C, the nozzle is closer than you set. Always tram with the nozzle heated to printing temperature to account for thermal expansion. Same for the bed — tram with the bed at printing temperature.
Mistake 3: Ignoring bed surface cleanliness.
Oils from your fingers destroy adhesion. Isopropyl alcohol (IPA) on a paper towel between prints removes skin oils. For PEI sheets, a light scuff with 0000 steel wool every 50 prints refreshes the surface texture. For glass, wash with dish soap and water when IPA stops working (oils build up in the microscopic pores that IPA doesn’t reach).
Mistake 4: Not verifying Z-offset after nozzle changes.
A new nozzle screws in to a slightly different depth than the old one — even identical brand nozzles. 0.1mm difference in seating depth means your old Z-offset is wrong. Always run a first-layer test after swapping nozzles. It takes 2 minutes and saves a failed print.
Mistake 5: Relying entirely on a bed probe without manual verification.
A probe measures the bed surface, but it doesn’t verify that the measurement is correct. A probe mount that loosens by 0.2mm introduces a 0.2mm offset in the mesh. Watch the first layer of every print for the first 30 seconds. The probe is a tool, not a substitute for your eyes.
⚠️ Safety Notice: 3D printers operate at high temperatures and involve electrical components. Always ensure thermal runaway protection is enabled. Keep flammable materials away from the heated bed. Some bed adhesives (hairspray, glue stick) produce fumes when heated — ensure adequate ventilation. Nozzles and beds reach temperatures that cause burns — let components cool before handling.
For a deeper dive into bed surfaces, our bed adhesion guide covers PEI vs glass vs BuildTak and surface preparation techniques. If you’re seeing inconsistent first layers across the bed even after tramming, our bed mesh leveling guide explains probe setup and mesh density optimization.
When your first layers are consistently perfect, try printing PETG drone parts — camera mounts, antenna holders, and GPS stands. PETG’s layer adhesion depends heavily on first-layer squish, and a dialed-in Z-offset is the difference between a mount that shatters on the first crash and one that survives a season. The uavmodel store stocks PETG specifically formulated for structural drone accessories.