Why Your Hatch Disappears or Turns Solid (and How to Fix the Scale)

Two symptoms, one root cause

Almost every hatch complaint I get asked about boils down to one of two things happening on screen: either the hatch you just applied is simply not there, the boundary highlights and confirms but nothing visible fills it, or the opposite happens and the whole area goes solid black, dense enough that you cannot make out any pattern lines at all. Both look like completely different problems and get treated that way, one as a missing render and the other as a corrupted fill, but in the overwhelming majority of cases they are the exact same root cause showing up at opposite ends of a scale: the hatch pattern's spacing is either far too large relative to your drawing units, so the repeating lines never actually cross your boundary at a visible interval, or far too small, so the lines are packed at a density your screen and printer cannot resolve and it reads as solid.

I want to be clear this is a genuinely common thing, not a sign you have done something unusually wrong, hatch scale is one of the more counterintuitive settings in AutoCAD because it interacts with drawing units, annotation scale, and viewport scale all at once, and getting any one of those out of alignment with the others produces exactly these two symptoms. This guide walks through diagnosing which one is happening and fixing it without guessing at random scale numbers until something looks right, which is honestly how most people end up solving it and it wastes a lot of time.

Symptom A: the hatch that seems to vanish

When a hatch appears completely blank after you apply it, the boundary is almost always fine, if AutoCAD let you complete the HATCH command without an error, it found a closed boundary, so the geometry is not usually the issue. What has actually happened is the pattern's line spacing, multiplied by whatever scale factor you set, is larger than the boundary itself, so between one repeating line and the next there is more empty space than your visible area contains, and you are looking at a fill that is technically there but sparse enough that zero lines happen to fall inside the boundary at the current view.

The fastest check is zooming out. If a hatch that looked empty at your current zoom suddenly shows a line or two once you zoom out several times, that confirms it, the scale is too large for the area you are filling, not broken. Reduce HPSCALE, either through HATCHEDIT or the Properties palette scale field, in increments, say cutting it in half repeatedly, and re-check after each change rather than guessing at a single number, because the right scale depends entirely on how large your boundary is relative to the pattern's base unit spacing.

Symptom B: the hatch that turns solid black

The opposite problem, a hatch rendering as an unbroken black mass, is the same mismatch running the other direction, the pattern's spacing multiplied by scale is now far smaller than your boundary, so instead of too few lines crossing it you have thousands of them packed closer together than your screen resolution or your printer's line weight can separate, and it visually merges into what looks like a solid fill even though it is technically still a line pattern underneath.

This is the single most common outcome of the metric versus imperial mismatch covered in the previous guide in this series, an inch based pattern applied inside a millimetre drawing effectively has all its spacing values read as roughly twenty five times smaller than intended, so lines that should sit an inch apart end up under a millimetre apart, dense enough to look solid at almost any zoom level. The fix is the same mechanism as symptom A but in reverse, increase HPSCALE substantially, often by a factor in the twenty to thirty range when it is a unit system mismatch specifically, and check the result at actual print scale rather than a zoomed in model space view, because model space zoom can be misleading either way.

Why hatch scale is relative to drawing units, not paper

The core thing to internalize is that HPSCALE multiplies the pattern's base spacing in real world drawing units, not in printed inches or millimetres on the sheet, because AutoCAD assumes, correctly, that you are drawing everything at true 1 to 1 real world size in model space and only scaling down for output at plot time. So a brick hatch pattern with a base spacing designed to represent roughly 200 to 230 millimetre long bricks, a realistic range for a standard building brick, needs its scale set relative to your drawing's real unit, not to how big you want it to look on an A3 sheet.

This is where a lot of people, reasonably, get tripped up, because they are thinking in terms of "how big should this look when I print it" and adjusting scale until it looks right on screen at whatever zoom they happen to be at, rather than thinking in terms of the actual real world spacing the pattern is meant to represent. Once you separate those two ideas, real world hatch scale versus how it will ultimately look at plot scale, on a viewport set to something like 1 to 50 or 1 to 100, the whole thing gets a lot less trial and error, and you can set a hatch scale once based on the material's real dimension and trust it across every viewport at every plotted scale.

On a drawing where the same wrong scale has been applied to several hatches at once, correcting each one individually through Properties can get tedious fast, and it is worth knowing AutoCAD lets you select multiple hatch objects together and adjust the scale field once across the whole selection, provided they all genuinely need the same correction, rather than reopening Properties separately for every single hatched area. This matters most right after discovering a unit system mismatch has affected an entire imported drawing, where dozens of hatches across many rooms or many site areas all came in at the same wrong density, since fixing that as one batch selection rather than one area at a time is the difference between a two minute correction and a genuinely tedious afternoon.

User defined versus predefined pattern type, another variable worth checking

One more variable that gets tangled up with HPSCALE and is worth separating out clearly, AutoCAD's hatch dialog lets you choose between a predefined pattern, one of the built in library patterns like ANSI31 or the standard brick and paving patterns referenced earlier in this guide, and a user defined pattern, which instead of pulling from a saved spacing definition, lets you set the line spacing and angle directly as simple values you type in yourself. A user defined hatch does not use HPSCALE the same way a predefined one does, its spacing is set directly rather than being a base spacing multiplied by a scale factor, so if you find yourself adjusting HPSCALE on a hatch and seeing absolutely no change in the result at all, checking whether that specific hatch was actually created as user defined rather than predefined is worth doing before assuming something else is broken, since scale simply does not apply the same way to that hatch type.

The practical implication, if a hatch was created as user defined with its spacing typed in directly, and it is coming out too dense or too sparse, the fix is adjusting that spacing value directly through HATCHEDIT rather than hunting for a scale setting that will not do anything, and it is a genuinely common source of confusion for someone used to always reaching for HPSCALE first regardless of pattern type. Confirming which type you are actually looking at, right there in the same Properties palette pattern section mentioned earlier, takes a few seconds and saves a fair bit of wasted troubleshooting on the wrong control.

Model space, paper space, and where annotation scale fits in

If you are placing hatches directly in a layout viewport rather than model space, or you have annotation scaling turned on for the hatch object, there is a second layer of scale sitting on top of HPSCALE, and mixing these up is where the confusion compounds. Annotation scale, when active on a hatch, adjusts the displayed pattern density based on which annotation scale is current for that viewport, so the same hatch object can look correctly spaced in a 1 to 50 viewport and completely wrong, too dense or too sparse, in a 1 to 100 viewport of the same drawing, if the annotation scale representations were not added for both.

The practical fix, and the one I default to on almost every project, is keeping hatches non annotative and drawn in model space at true units, then letting the viewport's own scale handle how large it appears on the sheet, exactly the same way you would treat a wall or a door. Reserve annotative hatch scaling for the specific case where the same drawing genuinely needs to be plotted at multiple different scales with the hatch looking correctly proportioned at each one, which is a narrower use case than most people reach for it as.

Fixing it: a worked example using a paver joint

Say you are hatching a courtyard paving area and want the joint lines to roughly match a standard paver unit, something typically in the 100 to 200 millimetre range depending on the paver, in a drawing set up in millimetre units. Select the hatch, open the Properties palette with Ctrl+1, find the Pattern section, and adjust the Scale field, starting from AutoCAD's default of 1 and working upward, checking the result at your actual intended plot scale rather than a random zoom level, since what looks right at a tight zoom can look wrong once plotted at 1 to 100.

If you want that same courtyard to genuinely show individual paver joints rather than an approximated fill, and the client or reviewer is going to zoom into that area on a presentation sheet, this is exactly the case where swapping to a real drawn block, paving-block-1 or one of the other numbered units in our paving category for instance, laid out unit by unit along the path, reads more honestly than tuning a hatch scale to approximate the same joint spacing. You are not fighting HPSCALE numbers at that point, you are placing real geometry at its actual size, which sidesteps the whole scale guessing game for that specific area.

Using QSELECT to find every affected hatch quickly

When a units mismatch or a scale problem has affected more than a couple of hatches scattered across a large imported drawing, manually clicking through the drawing looking for every affected fill is slow and genuinely easy to miss one or two instances of, particularly on a busy site plan with dozens of separate hatched regions. QSELECT, AutoCAD's quick select tool, lets you filter the entire drawing for every hatch object matching a specific property, a particular pattern name, a specific layer, or an object type generally, and select them all simultaneously regardless of where they sit in the drawing, which is exactly the tool for confirming you have actually caught every instance of the problem rather than just the ones you happened to notice while scrolling through the sheet.

Running a QSELECT filtered to hatch objects specifically, right after importing a drawing from an unfamiliar source or a different office's template, is a genuinely useful habit before doing any detailed review work on that file, since it surfaces the full scope of what you are dealing with immediately rather than discovering additional broken hatches piecemeal over the following hour of review.

When associativity breaks and mimics a scale problem

One more failure mode worth knowing before you assume every weird hatch is a scale issue: associative hatches, the default in modern AutoCAD, are supposed to update automatically when you stretch or edit their boundary, but that link can break silently after certain edits, a boundary object gets exploded, moved to a different layer state, or edited through a command the hatch does not track well, and the hatch is left referencing geometry that no longer matches.

What this looks like on screen is a hatch that appears to spill outside its boundary, or stop short of it, or occasionally vanish from that specific area while hatches elsewhere in the drawing look completely normal, and it is easy to mistake for a scale problem because the visual symptom, an odd looking fill, is similar. The tell is that HATCHEDIT will often show the pattern and scale values as completely normal and correct, nothing wrong there, while the boundary itself is the actual issue. If scale looks fine in the properties but the fill still looks wrong, re-create the hatch against the current boundary rather than continuing to adjust scale, since you are very likely chasing the wrong variable.

A print time sanity check worth running every time

- Zoom to the actual plotted scale in a viewport before judging a hatch, never trust model space zoom alone - Check HPSCALE relative to a real dimension you know, like a paver's actual joint spacing, rather than adjusting until it "looks right" - Confirm the drawing's base unit, millimetre or inch, matches the pattern's intended unit system before troubleshooting anything else - Re-select the hatch's boundary and re-create it if HATCHEDIT shows a normal scale but the fill still looks broken

Running through these four checks before a final print catches the overwhelming majority of last minute hatch surprises, and at the end of the day almost every one of them traces back to one of the three root causes above, scale mismatch, viewport confusion, or a quietly broken boundary association, rather than anything actually wrong with AutoCAD itself.