Why Does My 3D Print Have Random Pits or Holes in the Walls?

Random pits or small holes in a 3D print wall usually mean extrusion stopped or shrank for a fraction of a line. The common causes are moisture flashing into vapor at the nozzle, a weak restart after retraction, a partial nozzle restriction, intermittent spool or feeder drag, excessive flow demand, or a seam setting that removes too much material. The pattern is more useful than the hole count: pits stacked in one vertical line point toward the seam, pits beside travel starts point toward restart pressure, and scattered craters accompanied by popping point toward filament condition.

Before changing settings, confirm that the marks are depressions rather than raised blobs. Then inspect the exact sliced file with seam, travel, retraction, speed, and volumetric-flow views enabled. A short continuous-wall test is especially valuable: if the pits disappear when the nozzle prints one uninterrupted spiral wall, start-stop behavior is the stronger suspect. If they remain random during continuous extrusion, investigate moisture, flow stability, the nozzle, and the feed path.

Short answer: what should you check first?

1. Use raking light and touch. A true pit is recessed or missing material. A raised dot belongs in the blobs and zits guide.
2. Map the pattern. Mark whether pits form a vertical seam, follow corners, appear after travels, cluster at high-speed regions, or scatter randomly.
3. Listen during a supervised repeat. Popping or sizzling supports a moisture or contamination diagnosis. Clicking supports a flow-resistance or feeder diagnosis.
4. Check the slicer preview. Compare each pit height with seam position, travel starts, retractions, outer-wall speed, volumetric flow, thin-wall handling, and feature changes.
5. Print a continuous-wall coupon. Use a simple spiral or vase-style wall only as a diagnostic. If it prints cleanly, restarts and seams deserve attention before hardware.
6. Change one supported variable. Test spool condition, restart behavior, flow demand, or the feed path separately so the result identifies the cause.

Read the pit pattern before changing settings

1. First separate pits from blobs, seams, and layer cracks

A photograph can reverse the apparent shape of a small surface defect. Light from above may make a raised dot look like a crater, while a dark filament can hide a shallow gap. Let the part cool, place a light almost parallel to the wall, and rotate the part. A pit casts a shadow on the far side and usually catches a fingernail as a depression. A blob rises above the neighboring lines.

This distinction prevents the most common wrong fix. Increasing flow to fill a raised zit makes the wall more overfilled, while adding retraction to a missing restart can deepen the hole. If the marks are raised and scattered, use the blobs and zits guide. If they form one repeated zipper-like ridge or divot, use the seam-bump troubleshooting guide.

Also separate isolated pits from a continuous horizontal crack. A missing or weak line that wraps around the part suggests a longer feed interruption, layer-bond failure, or motion event. Random pinholes interrupt only short sections of individual extrusion paths. Keep the failed part and mark printer front, model front, and the print start so later comparisons remain meaningful.

2. A weak seam or restart can leave a small hole at every layer start

Most ordinary walls are not one continuous extrusion. The nozzle ends a perimeter, travels or changes feature, then starts the next line. Pressure inside the nozzle falls and rebuilds during that sequence. If the slicer retracts too far, unretracts too slowly, applies a negative restart amount, wipes too aggressively, or leaves a seam gap, the next line can begin short of material. The result is a recessed dot at or just after the seam.

Turn on seam and travel views for the exact G-code. If the holes stack where the seam markers sit, do not diagnose random moisture first. Print a short cylinder or box with the seam deliberately placed on an easy-to-inspect face. Then move the seam without changing flow, temperature, or retraction. A defect that moves with the commanded seam is strong evidence for start-stop behavior.

Use a continuous-wall test as a branch test, not as the final profile

A spiral or vase-style wall removes ordinary layer-start seams and most retractions. If that wall is clean with the same spool, nozzle, temperature, and approximate wall speed, the hotend can probably sustain continuous extrusion. Return to seam placement, retract/unretract behavior, wipe, coasting-like features, pressure advance, and travel strategy. If the spiral wall still has pits, the root cause is less likely to be the seam alone.

Do not copy a generic retraction distance from another printer. Direct-drive and Bowden systems, hotends, materials, temperatures, and travel lengths behave differently. Use the printer or slicer maker's calibration method and stay within supported ranges. Retraction should control ooze without pulling softened filament into a region where it swells, cools, or loses reliable grip.

3. Wet filament can create random craters when vapor disrupts the melt

Moisture absorbed by filament can flash into vapor in the hotend. A small bubble may expand at the nozzle, make a pop, and leave an irregular cavity or rough patch in the deposited line. Strong supporting evidence includes sizzling, fine bubbles in the extrusion, inconsistent gloss, increased stringing, a rougher top surface, or performance that worsened after the spool sat exposed. PETG, TPU, nylon, and some filled materials can show the problem readily, but even PLA can drift after poor storage.

One clean seam line is weak evidence for moisture. A razor-straight hole at every layer start should still be investigated as a restart problem. Moisture-related pits tend to be less orderly, although the exact appearance depends on wall speed, line width, polymer, and lighting.

When the evidence supports moisture, follow the filament maker's limits and the filament-drying guide. Compare the suspect spool with a known-dry reference while keeping the rest of the print unchanged. Do not assume a warm enclosure, sealed bag, or desiccant packet has actively dried the center of a wet spool. Storage slows reabsorption; it is not automatically a recovery cycle.

Contamination can imitate a moisture pop

Residue from a previous material, degraded polymer left hot for too long, dust, or foreign particles can disturb flow and leave a pit. The clue is often intermittent discoloration, rough extrusion, or a defect that persists with a known-dry spool. If material compatibility or purge volume is in doubt, use the printer and hotend maker's changeover procedure. Avoid burning filament out with uncontrolled heat or pushing metal tools into a powered hotend.

4. A partial clog can interrupt extrusion without stopping the print

A nozzle does not have to be fully blocked to create holes. A particle or degraded plug can move within the melt channel, temporarily reducing flow and then releasing. The part may show thin sections, rough lines, clicking, or intermittent recovery rather than one complete failure. Small nozzles, filled materials, dirty filament, and material changes can raise the risk, but a new nozzle is not immune to assembly or contamination problems.

Observe whether the defect appears during continuous extrusion and whether the extruder starts clicking at the same time. Inspect the purged strand only as supporting evidence; free-air extrusion uses less pressure than printing a fast wall and cannot prove full high-flow health. If the symptoms fit a restriction, use the nozzle-clog troubleshooting guide and the manufacturer's safe cleaning or replacement procedure.

Do not repeatedly raise temperature to force material through an unknown obstruction. Excess heat can degrade polymer, worsen ooze, damage temperature-sensitive hotend parts, or hide the problem briefly. Stop using the machine if temperature readings are unstable, heating is uncontrolled, wiring is damaged, or molten filament is leaking around the heater block.

5. Flow demand can outrun the hotend and leave holes in fast regions

Every combination of material, nozzle, layer height, line width, temperature, and hotend has a sustainable volumetric-flow range. When requested flow exceeds melt capacity, the extruder pressure rises while the deposited line becomes thin or intermittent. A slow first layer and small test cube may look normal, while a large wall develops pits only after acceleration reaches full speed.

Color the sliced model by speed and volumetric flow. Compare the first bad region with outer-wall speed, internal-feature transitions, and any flow cap. If pits cluster in the fastest long walls or the extruder clicks only there, reduce the demanding wall speed for a short coupon. A clean result supports a flow-limit diagnosis more strongly than increasing temperature and flow together.

Use a realistic validated flow cap. Raising the slicer's maximum flow number does not increase physical melt capacity. Higher nozzle temperature may expand the usable range, but it can also change gloss, dimensions, stringing, overhangs, polymer condition, and layer behavior. Confirm the full part after any temperature change rather than judging one cleaner wall.

A clean purge line does not prove high-flow wall performance

Extruding slowly into free air or laying down a wide purge line requires less sustained pressure than printing a long wall at production speed. Treat a smooth purge as evidence that material can leave the nozzle, not proof that the hotend can melt and deliver the requested volume continuously. The controlled wall coupon is the more relevant test.

6. Spool drag and feeder instability can create short missing sections

Intermittent resistance upstream of the hotend can remove material from an otherwise healthy extrusion path. A crossed winding, spool rubbing a wall, tight guide bend, dragging automatic-material-system path, worn tube, heavy side-mounted spool, or snag near the end of a roll can make feeder tension vary. The extruder may click, grind a notch, or simply deliver less material for a moment.

With the machine cold and safe, trace the filament path from spool to hotend. The spool should turn freely without overrunning, the filament should not rub a sharp edge, and tubes should follow the maker's bend and seating requirements. During a supervised print, watch whether pits correspond to spool rotation, carriage position, or a repeatable feed-path bend. Do not bypass covers, door switches, or routing safeguards to make the test easier.

If clicking is present, the timing matters. A click on the first layer suggests a nozzle that is too close; a click only on fast infill suggests flow demand; a click after enclosure heat builds suggests heat creep; a click at random spool positions suggests feed drag. The extruder-clicking guide provides that branch-by-branch sequence.

7. Retraction, wipe, coasting, and pressure advance can remove too much material

Several features try to manage nozzle pressure near line ends and travel moves. Retraction pulls filament back. Wipe moves the nozzle while pressure falls. Some slicers or firmware workflows reduce extrusion before the endpoint. Pressure or linear advance changes commanded feed around acceleration. Each can be useful, but overlapping or poorly calibrated compensation can create a divot before a travel or a starved restart afterward.

Inspect where the pit sits relative to the path. A hole immediately before travel suggests end-of-line removal or wipe behavior. A hole immediately after travel suggests weak restart pressure. A defect at sharp speed changes may involve pressure advance. Use the supported calibration for the exact printer, material, nozzle, and speed range. Change only one compensation family at a time.

Avoid treating random-pit repair as a contest to minimize retraction. Too little can create strings and residue; too much can create restart gaps, heat-creep risk, and filament damage. The goal is the smallest reliable intervention for the actual travel path, not an extreme setting copied from a profile screenshot.

8. Model and slicing transitions can look like missing extrusion

Some apparent holes are generated by geometry. A damaged mesh, extremely thin feature, embossed texture, small cavity, wall-thickness transition, or variable-width decision can make the perimeter step inward for part of a line. The print may be reproducing the sliced path accurately even though the original model looked smooth in shaded CAD view.

Scrub the layer preview at the exact pit height and zoom in far enough to see the external path. Compare feature type, line width, wall count, gap fill, seams, and tiny travel moves. If the path itself contains the notch, repair the model or change the relevant thin-wall strategy on a short test. Do not increase flow to fill a deliberate or slicer-generated recess; that changes every other line as well.

A controlled test sequence for random wall pits

1. Keep the failed part, exact project file, G-code, and spool together.
2. Let the part cool, mark printer orientation, and inspect it under fixed raking light.
3. Classify each mark as recessed, raised, cracked, discolored, or merely glossy.
4. Mark whether pits align vertically, follow corners, or scatter without a height pattern.
5. Open seam, travel, retraction, speed, volumetric-flow, line-width, and feature-type previews.
6. Compare the pit locations with layer starts, travel exits, high-flow walls, and geometry changes.
7. Run a supervised short repeat and listen for popping, sizzling, clicking, or grinding.
8. Print a continuous-wall coupon with the same spool, nozzle, temperature, and similar wall speed.
9. If the continuous wall is clean, test seam and restart behavior on a short ordinary-wall coupon.
10. If it still has pits, compare a known-dry spool and reduce only the highest wall-flow demand.
11. If clicking, thin extrusion, or contamination persists, inspect the feed path and follow the supported clog procedure.
12. Repeat the winning coupon twice, then verify the full part's dimensions, strength, seams, and finish.

Fixes that match the evidence

• Pits move with the seam: calibrate restart and seam behavior, reduce overlapping pressure-control features, or place the seam on a noncritical face.
• Pits disappear in a continuous wall: focus on retraction, wipe, travel, seam gap, and pressure advance rather than replacing the nozzle first.
• Pits are random with audible pops: validate the spool with a supported dry-then-store process and compare a known-dry reference.
• Pits appear with clicking or thin lines: identify whether resistance comes from the first-layer gap, requested flow, clog, heat creep, spool path, or feeder hardware.
• Pits cluster on fast long walls: lower that wall's flow demand or establish a realistic volumetric-flow limit.
• Pits repeat with spool position or carriage location: correct the verified drag, routing, tube, or cable condition using the printer maker's procedure.
• The preview contains the notch: repair the model or adjust the specific thin-wall/toolpath decision; do not compensate with global flow.
• Marks are actually raised: switch to blob, zit, or seam-bump diagnosis before changing extrusion quantity.

What not to change first

• Do not raise global flow from one wall photo. It can crowd top surfaces, swell dimensions, and worsen seams without fixing an intermittent gap.
• Do not add more retraction automatically. Excess retraction is one cause of weak restarts and can increase heat-creep or filament-damage risk.
• Do not dry every spool indefinitely. Look for popping, rough extrusion, exposure history, or a known-dry comparison.
• Do not replace the nozzle before checking the sliced path. A commanded seam gap or model notch will survive new hardware.
• Do not raise temperature beyond supported limits to hide clicking. Find the flow resistance and protect the hotend.
• Do not disable several pressure-control features at once. You will not know which interaction caused the pit.
• Do not sand a structural or sealing wall and call it repaired. Surface cleanup does not restore missing material or internal continuity.

Frequently asked questions

Are random holes in a 3D print always caused by wet filament?

No. Wet filament is likely when pits are irregular and accompanied by popping, bubbles, stringing, or unstable finish. A vertical line of holes is more consistent with seam or restart behavior, while holes in fast regions can indicate a flow limit.

Why do the holes appear only on the outer wall?

The outer wall may use a different speed, acceleration, line width, seam position, or print order. It is also the surface where tiny restart gaps are visible. Compare feature settings and flow demand instead of assuming the entire hotend is under-extruding.

Can too much retraction cause pits?

Yes. Excessive retraction can delay pressure recovery, damage the filament, or pull softened material into a cooler part of the hotend. The strongest clue is a pit immediately after travel that disappears in a continuous-wall test.

Can pressure advance cause small holes?

A badly mismatched value can contribute around acceleration and deceleration, especially when combined with wipe or restart compensation. Use the printer or firmware maker's calibration method and inspect whether defects track speed changes.

Should I increase nozzle temperature to fill the holes?

Only when evidence shows the material is not melting fast enough at the requested flow and the new temperature remains supported. Higher temperature does not repair a seam gap, wet spool, snagged feed path, or damaged mesh.

Will a new nozzle fix random pockmarks?

It can fix a damaged or persistently restricted nozzle, but it will not fix moisture, retraction, excessive wall speed, spool drag, or a sliced notch. Prove the restriction branch before replacing parts.

Next steps

Save the failed part, exact sliced project, spool identity, nozzle, temperatures, wall speed, flow cap, retraction settings, seam mode, and the result of the continuous-wall test. The common print-quality problems guide is the best next route when the marks occur with weak layers, rough tops, stringing, dimensional drift, or other symptoms that do not fit one branch.

For production parts, define whether pits are merely cosmetic or can compromise sealing, cleanliness, electrical spacing, fit, fatigue life, or customer-facing finish. Do not approve a process from one clean coupon; repeat it and inspect the critical faces under consistent light. If building and validating that process in-house costs more than the job justifies, the printer-versus-service guide provides the decision checkpoint. JC Print Farm is the relevant handoff when the file revision, material, quantity, critical dimensions, visible faces, and acceptance method are already defined.