CAD Layer Standards: How to Organise a Drawing So Others Can Use It

Layers are the part of CAD that decides whether a drawing can be used by anyone other than the person who drew it. A file where walls, furniture, dimensions and hatch all sit on the same layer works perfectly well until someone needs to turn the furniture off, plot at a different weight, or overlay the structural engineer’s file. At that point the drawing stops being a document and becomes a picture that has to be redrawn.

This guide covers what a layer standard does, how the two dominant naming systems work, what properties belong on a layer rather than an object, and what happens to all of this when a practice moves to BIM.

Why Layer Standards Exist

A drawing file is not a private document. It is exchanged with consultants, sent to contractors, imported into other drawings, referenced across a project for years, and often reopened by someone who has never spoken to its author. A layer standard is the agreement that makes all of that possible without a conversation.

Concretely, it delivers three things. Visibility control: the ability to switch categories of information on and off, so one base file can produce a general arrangement plan, a fire strategy plan and a furniture layout. Print control: consistent line weights across a whole set, because weight is assigned by layer rather than object by object. Exchange: a consultant who receives your file can find the walls without inspecting every entity, because the layer name tells them.

The absence of a standard costs more than it appears to. Time lost hunting for geometry, plots that come out with the wrong hierarchy, and coordination errors when two disciplines’ files cannot be meaningfully overlaid are all downstream of the same decision not to organise the file.

The AIA and US National CAD Standard Format

The dominant system in North America comes from the AIA CAD Layer Guidelines, absorbed into the US National CAD Standard. It builds a layer name from up to four parts, separated by hyphens.

Discipline designator comes first: A for architectural, S structural, C civil, L landscape, M mechanical, E electrical, P plumbing, F fire protection, I interiors, G general.

Major group is a four-character code for the building system: WALL, DOOR, GLAZ, FLOR, CLNG, ROOF, EQPM, FURN, ANNO for annotation.

Minor group, optional and also four characters, narrows it further: FULL for full-height, PRHT for partial-height, IDEN for identification tags, DIMS for dimensions, PATT for patterning.

Status, a single optional character, records the work’s phase: N new, E existing, D to be demolished, F future, T temporary, X not in contract.

The system’s real strength is that it sorts alphabetically into something useful. Every architectural layer groups together, every annotation layer groups within that, and a layer list becomes navigable rather than a wall of names.

ISO 13567 and UK Practice

Outside North America the equivalent reference is ISO 13567, which structures layer names into mandatory and optional fields covering agent or discipline, element, and presentation, with room for project-specific extensions. In the UK, BS 1192 established the layer conventions that most practices grew up with, and although information management has since moved to the ISO 19650 framework, the underlying discipline of structured, meaningful layer names carried straight across.

The systems differ in their codes but agree completely on the principle: a layer name is a structured field, not a free-text label. A-WALL-FULL can be parsed by a person or a script. walls new final v2 cannot be parsed by either.

What Belongs on the Layer, Not the Object

A layer standard only works if the drawing’s properties actually follow the layer. In practice that means setting colour, line type and line weight to ByLayer on every object, and controlling the appearance centrally.

Colour is usually the workhorse, because colour-dependent plot styles map each colour to a printed line width. This is why an AutoCAD drawing frequently looks garish on screen and prints as a clean grey hierarchy: the colours are carrying weight information, not aesthetic intent. Named plot styles achieve the same result without commandeering colour, and are cleaner in principle, though colour-dependent styles remain more common in practice.

Line type belongs on the layer too, so that hidden work is dashed and grid lines are chained without anyone having to remember. Line weight can be set on the layer directly, though many offices continue to resolve it at plot time through the plot style table instead. Either is defensible; mixing them within one set is not, and the resulting hierarchy is covered in the guide to architectural line weights.

The rule of thumb is that object-level overrides should be rare and deliberate. A drawing where individual entities carry their own colour and weight cannot be restyled, cannot be plotted consistently, and cannot be handed to anyone else with confidence.

What Happens to Layers in BIM

Revit has no layers, and this surprises people moving over from CAD. The equivalent structure is distributed across several different mechanisms.

Categories and subcategories replace the layer as the classification of what an element is: walls, doors, furniture, and beneath those, subcategories controlling parts of a family. Visibility and graphics are controlled per view through these, rather than by switching layers on and off globally.

Filters handle conditional display, showing all fire-rated walls in red, for instance, based on parameter values rather than on which layer something was drawn on. Worksets handle collaboration, dividing the model so multiple people can work on it. View templates then bundle these settings so they apply consistently.

Layers reappear at the boundary. When a Revit model is exported to DWG, every category is mapped to a layer name, and that mapping is configurable, including by loading a standards-compliant mapping file so the exported drawing arrives at the consultant already conforming to NCS. Getting this export setup right is what makes a BIM-based practice a good collaborator with CAD-based ones, and it is worth configuring once at template level rather than per export.

Common Mistakes

Drawing everything on layer 0. Layer 0 has a specific role in block definitions, where it allows a block to inherit properties from the layer it is inserted on. It is not a general-purpose drawing layer.

Free-text layer names. Names such as “new walls”, “walls2” and “FINAL walls” accumulate until nobody can tell which holds current geometry.

Object-level overrides. Setting colour or weight directly on entities defeats the entire system and makes consistent plotting impossible.

Importing without purging. Referencing another consultant’s file drags their entire layer structure into yours. Bind, purge and audit before you inherit somebody else’s mess.

Never deleting. Layers accumulate across a project. Purging empty layers periodically keeps the list navigable and the file smaller.

Frequently Asked Questions

What is the standard for CAD layer naming?

In North America, the AIA CAD Layer Guidelines as absorbed into the US National CAD Standard, which builds names as discipline, major group, minor group and status: A-WALL-FULL-N. Internationally, ISO 13567 defines an equivalent structured approach, and UK practice grew out of BS 1192. The codes differ but the principle is identical: a layer name is a structured field that can be parsed, not a free-text label.

What does A-WALL-FULL mean?

A is the architectural discipline, WALL is the major group for walls, and FULL is the minor group meaning full-height. So the layer holds architectural full-height walls. Adding a status character makes it more specific: A-WALL-FULL-N is new full-height walls, A-WALL-FULL-D is full-height walls to be demolished.

What are the status codes in NCS layer names?

A single character appended to the layer name: N for new, E for existing to remain, D for existing to be demolished, F for future work, T for temporary, and X for work not in contract. They matter most on refurbishment projects, where the same drawing has to distinguish what is staying, what is going and what is being added.

Should line weight be set by layer or by object?

By layer, essentially always. Set every object’s colour, line type and line weight to ByLayer and control appearance centrally, either through layer properties or through a plot style table. Object-level overrides make a drawing impossible to restyle or plot consistently, and they are the main reason a set that looked correct on one machine prints wrongly on another.

Why do AutoCAD drawings use so many colours?

Because in colour-dependent plot styles the colour is carrying line weight information rather than aesthetic intent. Each screen colour maps to a printed width in the CTB file, so a drawing that looks garish on screen plots as a clean grey hierarchy. Named plot styles do the same job without using colour, and are cleaner in principle, but colour-dependent styles remain widespread.

Does Revit have layers?

No. Revit classifies elements by category and subcategory, controls conditional display through filters, divides work through worksets, and bundles view settings through view templates. Layers only appear on export: when a model is exported to DWG, each category is mapped to a layer name, and that mapping can be configured to produce standards-compliant output for consultants working in CAD.

What is layer 0 for in AutoCAD?

Layer 0 exists so that geometry inside a block definition can inherit the properties of whatever layer the block is inserted on. Objects drawn on layer 0 within a block take on the colour, line type and weight of the host layer. That is its purpose; using it as a general drawing layer defeats the mechanism and leaves geometry outside the layer structure entirely.

How do I clean up a messy layer structure?

Audit the file, purge unused layers and blocks, then remap the remaining geometry onto a standard set rather than renaming layers in place, which tends to leave duplicates. Bind and purge any external references first, so you are not inheriting another consultant’s structure. Doing this once against an office template is far quicker than doing it repeatedly per drawing.

Where to Go From Here

Layer standards, line weights and hatching are three faces of the same discipline: deciding once how a drawing should look, then drawing without deciding again. For the printed hierarchy those layers ultimately produce, see the guide to architectural line weights. For the ready-made geometry that most often arrives with somebody else’s layer structure attached, see the roundup of free CAD block sources.