How to Get a Perfect First Layer Every Time

Most failed prints die in the first 0.2 mm. A perfect first layer is a smooth, slightly squished, fully fused sheet of plastic with no gaps, no ridges, and no corners lifting off the bed. Get it right and everything above it gets easier. This guide covers bed leveling, Z-offset calibration, temperatures, slicer settings, and the exact G-code to dial it all in.

Quick Takeaways

  • Z-offset is the biggest variable. Aim for a squish that gives a first layer roughly 0.2 mm tall and 0.4-0.5 mm wide for a 0.4 mm nozzle.
  • Clean the bed first. Isopropyl alcohol (90%+) or dish soap and warm water removes the oils that cause most adhesion failures.
  • Print the first layer slow: 15-25 mm/s with a 100-120% line width and a 0.2-0.3 mm layer height.
  • Calibrate hot. Set bed and nozzle to printing temperature before adjusting Z-offset, because the nozzle and bed expand when heated.
Symptom Likely Cause Fix
Gaps between lines Nozzle too high Lower Z-offset in -0.02 mm steps
Rough, ridged, translucent surface Nozzle too low Raise Z-offset in +0.02 mm steps
Corners lifting Dirty bed or low bed temp Clean bed, raise temp 5 °C
One side good, other side bad Bed out of tram Re-level or run mesh leveling

Why the First Layer Matters

The first layer is the foundation for every layer after it. Poor adhesion causes warping, spaghetti failures, and dimensional errors. A layer that is too thin creates elephant’s foot and clogs. A layer that is too thick may not stick at all.

You are balancing three forces:

  1. Adhesion: plastic must bond to the build surface.
  2. Flow: plastic must fill the gaps between lines without over-extruding.
  3. Flatness: the nozzle must travel a consistent distance from the bed across the entire surface.

Step 1: Prepare the Build Surface

Fingerprints, skin oil, and leftover adhesive create invisible barriers.

  • Smooth PEI / textured PEI: Wipe with 90%+ isopropyl alcohol (IPA). For stubborn grease, wash with warm water and dish soap, then dry completely.
  • Glass: Clean with IPA. Apply a thin layer of glue stick or hairspray if needed.
  • Painter’s tape / BuildTak: Replace when worn or lifted.

Do not touch the print area after cleaning. Hold the plate by the edges.

Surface and Temperature Reference

Material Bed Temp Nozzle Temp (1st Layer) Best Surface
PLA 55-65 °C 205-215 °C Textured or smooth PEI
PETG 70-85 °C 235-245 °C Smooth PEI (use glue stick as release agent)
ABS/ASA 95-110 °C 245-260 °C PEI, enclosure required
TPU 40-60 °C 220-230 °C PEI, slow speed
Nylon 70-90 °C 255-270 °C Glue stick on PEI or glass

Step 2: Mechanically Level and Tram the Bed

Software compensation cannot fix a badly warped or tilted bed. Start mechanical.

Manual leveling with paper

  1. Heat the bed and nozzle to printing temperature.
  2. Home all axes.
  3. Disable steppers or move the nozzle to each corner.
  4. Slide a sheet of standard printer paper (about 0.08-0.10 mm) under the nozzle.
  5. Turn the leveling knob until you feel light drag on the paper.
  6. Repeat at all four corners and the center. Go around twice, since adjusting one corner shifts the others.

Auto bed leveling (ABL)

Probes such as BLTouch, CRTouch, inductive probes, and load-cell sensors measure the bed at many points and compensate in software. ABL still depends on a properly trammed bed. Large height differences (more than about 0.5 mm across the plate) should be fixed mechanically first.

Marlin: enable bilinear mesh and save it:

G28            ; Home all axes
G29            ; Run auto bed leveling probe routine
M500           ; Save settings to EEPROM

Klipper: run a bed mesh calibration:

G28
BED_MESH_CALIBRATE
SAVE_CONFIG

To load the mesh automatically at print start, add this to your START_PRINT macro:

BED_MESH_PROFILE LOAD=default

Step 3: Dial in Z-Offset

Z-offset is the distance between the nozzle tip and bed when the printer thinks Z = 0. This is the single most important setting for first layer quality.

Using live Z-adjust (Babystepping)

  1. Preheat to material temperature and wait 5-10 minutes for thermal expansion to settle.
  2. Start a first layer test print (see Step 5).
  3. Watch the first few lines and adjust in 0.02 mm increments while printing.
  4. Save the final value once lines fuse into a smooth surface.

Marlin:

M851 Z-1.20    ; Set probe Z-offset (example value, yours will differ)
M500           ; Save to EEPROM

Klipper:

SET_GCODE_OFFSET Z_ADJUST=-0.02 MOVE=1    ; Lower nozzle 0.02 mm live
Z_OFFSET_APPLY_PROBE                       ; Apply to probe config
SAVE_CONFIG

Reading the squish

What You See Diagnosis Action
Round, separated lines with visible gaps Too high Lower nozzle
Smooth, flat, lines fully merged Perfect Save value
Ridged, wavy, translucent, or nozzle scraping Too low Raise nozzle
Rough texture with elephant’s foot on later layers Too low Raise nozzle slightly

A reliable check: the finished first layer should be flat to the touch with no raised seams between lines, and you should not be able to see individual extrusion lines on a solid fill.

Step 4: Optimize Slicer Settings

Your slicer controls how the first layer is deposited. These settings work well in Cura, PrusaSlicer, OrcaSlicer, and Bambu Studio.

Setting Recommended Value Why
Initial layer height 0.2-0.3 mm (0.4 mm nozzle) Thicker layers tolerate bed imperfections
Initial layer line width 100-120% of nozzle diameter Wider lines increase contact area
Initial layer speed 15-25 mm/s Gives plastic time to bond
Initial layer flow 100% (up to 105% if gappy) Fills gaps without bulging
Initial layer fan speed 0% Cooling hurts adhesion
Initial layer travel speed 100-150 mm/s Reduces oozing, no effect on adhesion
Bed temp (initial) Material spec +0-5 °C Slightly hotter bed improves stick
Brim/skirt Skirt 3-5 loops or brim 5-8 mm Skirt primes nozzle, brim prevents corners lifting

Add a skirt to prime the nozzle

A skirt purges old pressure and shows you how the first layer will lay down before the actual part begins. Use it as a free live preview. Stop the print if the skirt looks wrong.

Purge line macro (Klipper)

[gcode_macro START_PRINT]
gcode:
    G28
    BED_MESH_PROFILE LOAD=default
    G92 E0
    G1 X10 Y10 Z0.3 F5000
    G1 X10 Y150 E15 F1200   ; Prime line, 140 mm long
    G92 E0

Step 5: Run a First Layer Test

Do not tune on a real print. Use a dedicated test that covers the whole bed.

Option A: Single-layer square grid

Print a one-layer-thick square (for example 150 x 150 mm, 0.2 mm tall) at the center of the bed. Check the middle and all four corners.

Option B: Multi-square pattern

Slice five 30 x 30 mm squares at the four corners and center, each 0.2 mm tall. This reveals whether the bed is flat across the plate.

Option C: Spiral vase test

Print a single-layer concentric spiral to see raw extrusion consistency.

What to look for:

  • Edges should be consistent in width and fused to neighbors
  • No corners curling or peeling
  • Same quality in all five zones
  • Peels cleanly in one piece without tearing

Common Problems and Fixes

Problem Cause Fix
Lines won’t stick, peel off easily Dirty bed, nozzle too high, bed too cold Clean with IPA, lower Z by 0.02 mm, raise bed 5 °C
Elephant’s foot (bulging base) Nozzle too low, bed too hot Raise Z, lower bed temp 5-10 °C
Blobs or missing lines at start Nozzle not primed, oozing Add skirt or purge line, retract before travel
Corners lift on large parts Warping from cooling Add brim, use enclosure, raise bed temp
Nozzle scrapes bed or clicks Z-offset too low, extruder skipping Raise Z by 0.05 mm and retest
Good in center, bad on edges Bed not flat or tram off Run mesh leveling, check for warped plate
Stringy first layer Wet filament or high temp Dry filament 4-6 hours at 45-55 °C (PLA)

Practical Examples and Real-World Use Cases

Large flat parts (enclosures, trays, brackets)

A 250 mm-wide electronics enclosure base exposes every bed flatness issue. Run a full mesh, use a 0.3 mm initial layer height for forgiveness, and add an 8 mm brim to stop corner lift.

Functional parts with tight tolerances

Parts like press-fit gears or bearing housings need accurate first-layer dimensions. Over-squish causes elephant’s foot that shrinks hole diameters by 0.1-0.3 mm. Dial Z-offset precisely, or add 0.2 mm of first-layer horizontal expansion compensation (a negative value, for example -0.1 mm, in most slicers).

Print farms and repeat production

Production environments cut variance by standardizing: same bed surface, same material, same 0.2 mm Z-offset test, same pre-heat time. Many farms add a 5-minute heat soak before each job to stabilize bed geometry.

Calibrating after a hardware change

Re-tune Z-offset after any of these: nozzle swap, build plate swap, hotend repair, or firmware update. A new nozzle can change effective height by 0.05-0.15 mm.

Maintenance Habits That Keep the First Layer Consistent

  • Clean the plate before every print with IPA, or wash weekly with soap and water
  • Re-check Z-offset every 20-30 prints, or whenever quality drops
  • Inspect the nozzle for filament buildup (cold pull or replace every 3-6 months)
  • Check belt tension and frame rigidity quarterly, since a loose gantry causes uneven first layers
  • Replace worn plates (PEI coatings typically last several hundred prints with proper care)

FAQ

Why does my first layer look good in the center but not the edges?

The bed is not flat or not trammed. Run mesh bed leveling (G29 in Marlin, BED_MESH_CALIBRATE in Klipper). If the variance exceeds about 0.5 mm, shim the bed or replace a warped plate.

What is the ideal Z-offset for a 0.4 mm nozzle?

It depends on the printer, probe, and plate, so no universal number exists. Target a squish where a 0.2 mm first layer lays down about 0.4-0.5 mm wide, with no gaps and no ridges. Adjust in 0.02 mm steps until lines merge into a smooth surface.

Should I print the first layer fast or slow?

Slow. Use 15-25 mm/s. Slower speed gives molten plastic more time to wet the surface and bond, which is the main factor in adhesion.

Should I turn off the cooling fan for the first layer?

Yes. Set the fan to 0% for the first layer (and often the first 2-3 layers for PETG and ABS). Cooling the plastic before it bonds causes poor adhesion and warping.

Hot this week

How Much VRAM Do You Need? Model Size, Quantization, and Memory Explained

Calculate exact VRAM for any LLM: weights, quantization, KV cache, and GPU picks. Use our formulas and Python calculator to size your build today.

One-Hot vs Label vs Target Encoding: Which Should You Use?

Compare one-hot, label, and target encoding with code, a decision table, and leakage fixes. Pick the right categorical encoder for your model today.

Personal AI Agents: What They Can Do Today and What’s Still Hype

Which personal AI agents actually work in 2026? See benchmarks, security risks and a safe starter setup. Read the guide before granting access.

Handling Missing Data: Imputation Techniques That Actually Work

Learn the methods that hold up in production and the evaluation habits that expose the ones that don’t. Read the full guide now.

Streamlit and Gradio: Ship an AI Demo in 30 Minutes

Build and deploy an AI chat demo with Streamlit and Gradio in 30 minutes. Copy-paste Claude API code, compare both tools, and ship today.

Topics

How Much VRAM Do You Need? Model Size, Quantization, and Memory Explained

Calculate exact VRAM for any LLM: weights, quantization, KV cache, and GPU picks. Use our formulas and Python calculator to size your build today.

One-Hot vs Label vs Target Encoding: Which Should You Use?

Compare one-hot, label, and target encoding with code, a decision table, and leakage fixes. Pick the right categorical encoder for your model today.

Personal AI Agents: What They Can Do Today and What’s Still Hype

Which personal AI agents actually work in 2026? See benchmarks, security risks and a safe starter setup. Read the guide before granting access.

Handling Missing Data: Imputation Techniques That Actually Work

Learn the methods that hold up in production and the evaluation habits that expose the ones that don’t. Read the full guide now.

Streamlit and Gradio: Ship an AI Demo in 30 Minutes

Build and deploy an AI chat demo with Streamlit and Gradio in 30 minutes. Copy-paste Claude API code, compare both tools, and ship today.

Elegoo Centauri Carbon Review: Can a Budget Enclosed CoreXY Really Compete?

Our Elegoo Centauri Carbon review covers speed, print quality, materials, and flaws. See if this $299-class enclosed CoreXY is your next printer.

Best Desk Setup Accessories to 3D Print: Headphone Stands, Monitor Risers, and More

Print a headphone stand, monitor riser, cable organizer, and more. Get materials, settings, and tested designs. Start upgrading your desk today.

How Much Does 3D Printing Really Cost? Printer, Filament, Power, and Hidden Extras

Printer, filament, electricity, and hidden costs of 3D printing, with real per-print math and a cost formula. See your true budget before you buy.

Related Articles

Popular Categories