How To Calibrate Creality Ender 3

Calibrating an Ender 3 means working through seven checks in a fixed order: frame and mechanics first, then bed levelling, then extruder E-steps, then flow rate, then temperature (PID tuning), then axis steps, and finally test prints. The order matters because each step depends on the one before it — no point tuning flow on a printer whose extruder feeds the wrong amount of filament, and no point levelling a bed on a frame that is not square. A full calibration takes about an hour, and most steps only need repeating when hardware changes.

The 7 calibration steps:

1. Check the frame — screws, gantry squareness, eccentric nuts, belts
2. Level the bed
3. Calibrate E-steps (extruder steps per mm)
4. Set the flow rate (extrusion multiplier)
5. PID tune the hotend (and bed)
6. Check X/Y/Z axis steps with a printed cube
7. Run calibration test prints and read the results

This guide applies to the Ender 3, Ender 3 Pro, Ender 3 V2, Ender 3 S1 and Ender 3 V3 SE. The steps are the same throughout, and the model-differences section at the end explains where the newer machines differ — including what the V3 SE’s auto-calibration does and does not do.

Tools You Need

You do not need much:

• Digital calipers — essential for E-steps, flow and cube measurements; a cheap set is accurate enough.
• Paper (about 0.1 mm) or a 0.1 mm feeler gauge — for manual bed levelling.
• A fine marker and a ruler — for marking filament at 120 mm during the E-steps test.
• Your slicer (Cura or Creality Print) plus a way to send G-code — the LCD menus work, but a USB terminal (Pronterface, OctoPrint) makes the E-steps readout and PID tuning much easier.
• Filament — the one you print with most; PLA is fine for all of this.

Step 1: Check the Frame, Eccentric Nuts and Belts

Calibration starts with the hardware — software cannot compensate for a wobbly frame.

Tighten and check the screws. Go over the frame, uprights and gantry screws with the supplied Allen keys — snug, not forced. Do not miss the Z lead screw nut and the coupler grub screws on the motor shaft: a loose coupler makes Z movement inconsistent whatever your settings say.

Check the gantry is square. The X gantry should sit parallel to the bed. If one side rides higher — common on single-lead-screw Ender 3s — prints lean and levelling never stays put. Re-seat the gantry brackets so both sides move together without sagging.

Adjust the eccentric nuts. Each wheel carriage has one eccentric nut per three-wheel group. Turn it until the carriage has no play but still moves smoothly. Too loose and it rocks; too tight and the wheels flat-spot. The bed carriage is most often out — if the bed wobbles by hand, fix that before levelling.

Set belt tension. X and Y belts should twang lightly when plucked, with no slack, but not guitar-string tight — over-tensioning wears bearings. Ripples (ghosting) or slight layer shifts on one axis usually mean belt tension. Our Creality Ender 3 belt tensioner guide goes deeper if yours will not stay put.

Step 2: Level the Bed

Bed levelling (tramming) makes the nozzle-to-bed gap identical at every corner, which is what gives you a consistent first layer.

Manual method, step by step:

1. Heat the bed to printing temperature (60 °C for PLA) and the nozzle to about 150–200 °C. Metal expands when hot — level hot, not cold.
2. Disable the steppers (LCD: Disable steppers, or M84) so you can move the head and bed by hand.
3. Move the nozzle over the first corner, roughly above its levelling knob.
4. Slide the paper (or feeler gauge) between nozzle and bed, and turn the knob until the paper drags lightly — consistent friction, not a pinch and not free movement.
5. Repeat at all four corners in an X pattern (front-left → back-right → front-right → back-left), then go around again — adjusting one corner moves the opposite one. Two or three rounds is normal.
6. Finish at the centre. If it feels noticeably tighter or looser than the corners, the bed itself is slightly warped — very common on Ender 3s.

If you have a BLTouch or CR Touch (V2, S1, upgraded originals): the probe compensates for small bed warps, but it does not replace manual levelling. Level manually first, then set the Z-offset so the first layer squishes correctly.

First-layer test: print five large squares (one per corner, one centre). Each should have smooth, fused lines with no gaps and no ridges. Gaps mean the nozzle is too far; rough, ridged lines mean too close. Our guide to Ender 3 Pro bed leveling covers this test and live adjustment in more detail.

Step 3: Calibrate E-Steps (Extruder Steps per mm)

E-steps tell the extruder motor how far to turn to push exactly 1 mm of filament. If this is wrong, every print is over- or under-extruded and no slicer setting fixes it properly. It is the most skipped — and most valuable — calibration on an Ender 3.

The method:

1. Load filament and heat the hotend to printing temperature (200 °C for PLA). Never extrude cold.
2. Measure 120 mm of filament up from where it enters the extruder and mark it with the pen.
3. Extrude 100 mm slowly (LCD, or G1 E100 F100 from a terminal).
4. Measure from the extruder entry to your mark — what remains tells you how much actually moved.

The formula:

New E-steps = Current E-steps × (100 ÷ actually extruded mm)

Actually extruded mm = 120 − remaining length

Worked example: Current E-steps are the stock 93. You marked 120 mm, extruded 100 mm, and 26 mm remains. Actually extruded = 120 − 26 = 94 mm — the extruder under-fed by 6 mm.

New E-steps = 93 × (100 ÷ 94) = 93 × 1.0638 = 98.9

Applying it: Read the current value with M92, set the new one with M92 E98.9, then save with M500 — otherwise the printer forgets it on the next power cycle. On stock firmware you can also enter it via the LCD (Control → Motion → Steps/mm → Esteps/mm, then Store Settings). Run the test again to confirm you now land within about 1 mm of 100 mm. Repeat only when you change the extruder, its gear or the motor.

Step 4: Set the Flow Rate (Extrusion Multiplier)

E-steps make the extruder mechanically accurate. Flow rate (extrusion multiplier in Cura) fine-tunes how much plastic actually lands in the print, and varies slightly by filament.

Single-wall cube method:

1. Print a hollow 20 mm cube in spiralize/vase mode (or 0 top layers, 0% infill, one wall) at a known line width — say 0.4 mm.
2. Measure the wall thickness with calipers in several places per side, away from corners, and average the results.
3. Calculate:

New flow % = Current flow % × (target wall ÷ measured wall)

Example: Target wall 0.4 mm, measured average 0.44 mm, current flow 100% → new flow = 100 × (0.4 ÷ 0.44) = 91%.

Set the result in your slicer profile, not the printer — flow is filament-specific, and a value tuned for one PLA can be a percent or two off for PETG. Our Ender 3 Pro Cura settings guide shows where flow and the other key settings live.

Step 5: PID Tune the Hotend and Bed

PID tuning teaches the printer how its heaters behave, so temperatures hold within a degree or two instead of swinging ±5–10 °C — swings that cause visible banding and inconsistent extrusion.

From a terminal, run the hotend autotune at your usual printing temperature:

M303 E0 S205 C10

E0 is the hotend, S205 the target temperature (use your real printing temp), C10 ten cycles — more cycles, steadier result. The printer heats and cools repeatedly, then reports new Kp, Ki and Kd values. Apply them with:

M301 P<Kp> I<Ki> D<Kd>
M500

M301 sets the reported values, M500 saves them. For the bed (if your firmware has bed PID enabled): M303 E-1 S60 C10, then M304 and M500. Many stock boards run the bed in bang-bang mode, where bed autotune will not work — that is normal, not a fault. Re-tune after changing the hotend, heater, thermistor or sock.

Step 6: Check X/Y/Z Steps with a Calibration Cube

Only after E-steps and flow are correct, print a 20 mm cube and measure it once cooled, mid-axis — not at the base, where first-layer flare distorts the reading.

• X and Y should measure within about ±0.1–0.2 mm of 20 mm. If an axis is consistently off, adjust with the E-steps formula: new steps = current steps × (20 ÷ measured). This is rarely needed on a healthy Ender 3 — check belt tension first, because a loose belt mimics a steps error.
• Z: treat with caution. Z errors are almost never a steps problem — they come from the coupler, lead-screw binding or first-layer squish. Changing Z steps usually makes real prints worse; only touch it if the error is large, repeatable and the mechanics are proven good.

Step 7: Calibration Test Prints — and What They Tell You

Calibration is confirmed by prints, not numbers alone. Run these three, in order:

Temperature tower — the same feature printed across a range of temperatures. Pick the temperature with the cleanest bridges and least stringing as that filament’s default. Stringy at every temperature points to retraction or damp filament, not calibration.

Retraction test — varies retraction distance or speed per section. Choose the lowest retraction that prints clean; excessive retraction on the Bowden setup causes clogs and gaps after travel moves.

First-layer test, again — re-run the five-square pattern from Step 2 at your final settings. Frame, bed, extruder, flow and temperature all show up in this one layer; it is your end-to-end check.

Finally, save a copy of all your values (M503 prints them) — after a firmware update or EEPROM reset you can restore them in minutes.

Ender 3 vs Pro vs V2 vs S1 vs V3 SE: What Changes

Model	Levelling	E-steps/flow access	Notes
Ender 3 (original)	Fully manual	LCD menus or terminal	All 7 steps apply exactly as written
Ender 3 Pro	Fully manual	LCD menus or terminal	Same process; wider Y extrusion and better PSU change nothing
Ender 3 V2	Manual + optional CR Touch	Colour LCD menus or terminal	Probe compensates for warp; still level manually and set Z-offset
Ender 3 S1	CR Touch auto-levelling fitted	On-screen menus or terminal	Direct drive — E-steps value differs from Bowden models, method identical
Ender 3 V3 SE	Automatic calibration	Mostly automatic, Z-offset adjustable	See below

Ender 3 V3 SE calibration: on or off? Leave it on. On the V3 SE (and V3), auto-calibration is a routine you run from the menu, not a replacement for judgement: it probes the bed with the CR Touch and builds a mesh, measures the Z-offset automatically with its strain-gauge sensor, and sets input-shaping values — which genuinely covers Steps 2 and 6 for most users. What it does not guarantee is a perfect first layer on your plate and filament: if the first layer looks too squished or too loose afterwards, fine-tune the Z-offset manually in 0.05 mm steps. That adjustment on top of the automatic mesh is normal, not a fault. E-steps and flow are factory-set for its direct-drive extruder and rarely need touching unless you change extruder hardware.

Thinking about upgrading? See our best Creality 3D printer reviews.

Frequently Asked Questions

How do I calibrate an Ender 3 Pro — is it different from the Ender 3?

No — identical. The Pro shares the same frame, extruder and hotend, so all seven steps apply the same way, with the same default E-steps of 93.

How often should I calibrate my Ender 3?

Not before every print. Re-check levelling when the first layer degrades, E-steps only after extruder changes, and PID after hotend changes. A full pass every few months is enough for most owners.

What is an Ender 3 calibration print?

A small test model that exposes one setting: a 20 mm cube for dimensions, a temperature tower for temperature, a retraction test for stringing, a first-layer pattern for levelling. Each isolates one variable — read the result, adjust the matching setting.

I calibrated everything and prints still fail — what now?

Work backwards by symptom. A peeling first layer means levelling, Z-offset or a dirty bed — not flow. Ringing means belts or frame (Step 1). Under-extrusion with extruder clicking means a partial clog or a cracked plastic extruder arm, a known weak point. Calibration makes a healthy printer accurate; it cannot fix a failing part.

Does the Ender 3 V2 calibrate the same way?

Yes — same mechanics, same seven steps. The differences are convenience ones: a colour screen, built-in belt tensioners, and an optional CR Touch that compensates for bed warp after manual levelling.

Should I change my Z steps/mm to fix a cube that measures short?

Almost certainly not. A short Z measurement is usually first-layer squish, a loose coupler or lead-screw binding. Fix the mechanical cause; change Z steps only if the error is large, repeatable over the full height and the mechanics check out.