First Layer Not Sticking? 4 Causes, Ranked by Odds

Almost every guide to first-layer adhesion lists the same six or eight causes in no particular order. That's not useful, because they are emphatically not equally likely.
These four are ranked by how often they turn out to be the actual problem — not by how sophisticated they sound. Work down the list.
The 5-minute test
Before reading further, do this. It tells you which of the four applies.
Print a small 30 × 30 mm square, single layer, and watch it happen.
| What you see | Which cause |
|---|---|
| Filament drags and curls onto the nozzle | Z-offset too high, or surface contaminated |
| Sticks briefly, then peels from one corner | Bed temperature too low, or cooling too early |
| Sticks in some places, not others | Surface contamination or a warped bed |
| Comes out squashed and transparent, nozzle scraping | Z-offset too low |
| Sticks perfectly, fails on a later layer | None of these — it's not a first-layer problem |
That last row saves more time than the other four combined. A print that fails at layer 40 is not an adhesion problem.
1. A contaminated surface (roughly half of all cases)
This is the most common cause by a wide margin, and it's the one people skip because it feels too obvious.
The contamination is almost never what you think. It's not usually hand oil directly from handling — it's the residue from a paper towel, a cloth that's been washed with fabric softener, a build plate sprayed with a "bed adhesive" product weeks ago, or the air in a kitchen.
What it looks like: adhesion is fine for weeks, then degrades across the whole plate rather than in one spot.
The fix: warm water and plain dish soap, fingertips, a full minute, then rinse until water sheets evenly rather than beading. IPA is a between-print wipe, not a cure — and 70 % IPA can leave a residue that makes things worse.
Don't use: acetone, metal scrapers, or soap with moisturisers.
→ Full method: How to Clean a Textured PEI Plate
2. Z-offset (and the direction is usually "too high")
Z-offset is the second most common cause, and the error is usually in one direction: the nozzle is too far from the bed.
"A little too high" produces the classic symptom — the bead is round rather than flattened, so it only touches the plate along a thin line. It might even survive for a few layers before letting go.
How to read your first layer:
- Bead is round, sits on top → too high. Lower by 0.02 mm steps.
- Bead is flat, wide, slightly squashed → correct.
- Bead is transparent, nozzle dragging → too low. Raise by 0.02 mm.
- Bead has gaps or the line breaks up → too high, or flow is low.
Change in steps of 0.02 mm, not 0.1. A tenth of a millimetre is a large move and will overshoot the correct band entirely.
One common trap: if you have a BLTouch or similar probe, don't assume that removes the need for offset tuning. The probe measures the bed; the offset tells the printer where zero actually is. They're separate corrections.
3. Temperature and cooling
Third on the list, and most often a combined problem rather than a single fault.
Three things interact on the first layer:
- Bed temperature — PLA wants 55–65 °C, PETG 70–80 °C. Below the low end, adhesion drops off quickly.
- First-layer print speed — typically 20–25 mm/s. Fast first layers are a common cause of peeling.
- Part cooling — many profiles ramp the fan up from layer 2. On large flat parts that's enough to peel a corner. Keep cooling off for the first two or three layers.
There's a fourth factor that gets overlooked: ambient temperature and draughts. A printer in a garage in winter, or next to an air conditioner, will fail on large parts no matter how well everything is tuned. If a print fails consistently on one side, check whether something is blowing on it.
4. A warped or damaged surface
Last, because it's the least common — and because you can only conclude it after ruling out the first three.
Warped bed: the symptom is localised failures that always happen in the same region. A 0.1 mm dip across a 235 mm bed is enough. Check with a straight edge across several points, or watch the first-layer bead height as the nozzle crosses the bed.
Damaged surface: the textured side of a PEI plate wears smooth over time. The tell is that the surface looks different in the failing area — smoother, slightly shiny. If a water test shows the plate is clean and prints still fail in one region, the texture is gone there.
Rule it out properly: clean the plate twice, check Z-offset, check temperatures. Only then blame the surface.
What doesn't belong on this list
A few things get recommended constantly and rarely help:
- Glue stick on textured PEI. It fills the texture and reduces grip. It's a release agent for smooth surfaces where prints stick too well.
- Hairspray. Same problem, plus an uneven film you'll have to clean.
- Raising bed temperature to 80 °C for PLA. It buys a print and warps the part later.
- Brim tape. Sometimes necessary for small contact areas, but if you need it on a 30 × 30 mm square, the cause is one of the four above.
The order to work in
- Clean the plate properly. One minute, soap, water, done.
- Print the 5-minute test square.
- Adjust Z-offset in 0.02 mm steps.
- Check bed temp and first-layer speed.
- Only then look at the surface.
That order isn't arbitrary — it's sorted by how often each step actually solves the problem, with the cheapest checks first.
FAQ
Why does my first layer stick then peel off? Usually temperature: either the bed is too cool, or cooling ramps up too early. Try holding the fan off for the first three layers and raising the bed by 5 °C.
Does a lower Z-offset always fix adhesion? No. If the surface is contaminated or the bed is warped, lowering Z just makes the nozzle scrape while the part still lifts. Rule out contamination first.
Why did my first layer work for months and then stop? That's almost always surface contamination, not a settings change. Something in your workflow changed — a different cloth, a new roll of paper towels, a different soap.
Is 70 % isopropyl alcohol bad for PEI? It can leave a residue. 90 %+ is better, but soap and water removes baked-on grease that IPA can't touch.
Should I use a leveling probe instead of adjusting Z-offset? No — they do different jobs. The probe maps the bed; Z-offset tells the printer where zero is. You need both.
Can a warped bed be fixed? Sometimes — shimming or a glass plate over the low area. If the warp is small and consistent, manually adjusting the mesh in that region works.
内链建议(枢纽页,链出 4 条)
- → A1:PEI 清洗(第 1 原因)
- → A3:喷嘴堵塞(排除「以为是首层问题」的情况)
- → A5:硬化钢喷嘴(挤出量异常时)
- → 兼容性对照表(床温与尺寸)
CTA(文末) If you've cleaned the plate properly and it still lets go in the same spot every time, the surface is worn — a [replacement plate] costs less than the filament you'd lose testing.
Reddit 版本(发 r/3Dprinting)
标题 The four causes of first-layer failure, ranked by how often they're actually the problem
正文 看过的教程都把六个八个原因平铺列出来,但它们出现的概率差很多。按实际概率排:
- 板子脏了(大概占一半)。而且不是手上的油,是纸巾、抹布、之前喷的胶水残渣。温水+洗洁精洗一分钟,冲水看是不是均匀铺开。
- Z-offset 偏高。调步长用 0.02mm,别用 0.1mm,一步跳过头。
- 床温 / 首层速度 / 冷却。大平面件首层开后冷却很容易翘边,前三层先别开。
- 板面变形或磨损。这个要等前三个都排除了才能下结论。
不在这张单子上的:纹理板上涂胶(填平纹理反而更难粘)、喷发胶、把 PLA 床温拉到 80。
还有一个省时间的判断:如果打印是在第 40 层才失败,那就不是首层问题,别在这条路上浪费时间。
发布检查
- [ ] 标题 ≤ 60 字符
- [ ] 首段 120 字内出现目标词
- [ ] 「5 分钟测试」表格必须前置,且配图
- [ ] 四条内链全部自然嵌在正文,不堆在文末
- [ ] FAQ 上 FAQPage 结构化数据
- [ ] 这篇是枢纽页:发布后从 A1/A3/A4/A5 各回链一次