Random Dungeon Generation vs. Hand-Crafted Design: Which Approach Works Best?

When you need a dungeon, you can start in two very different places.

You can decide:

"This fortress needs three entrances, a flooded lower level, two competing factions, and a hidden route to the command chamber."

Or you can roll on a table or use a generator and discover:

"The eastern corridor turns south, opens into an octagonal chamber, and contains a shrine guarded by something unexpected."

Both approaches can produce excellent play.

Both can also produce bad dungeons.

The important question is not:

Which method is better?

It is:

Which decisions should be intentional, and which ones would benefit from surprise?

Random dungeon generation is useful when you want prompts, uncertainty, speed, or results you would not have chosen yourself.

Handcrafted design is useful when specific geography, information, themes, relationships, or campaign consequences need deliberate attention.

Most Dungeon Masters can benefit from learning both.

And in practice, some of the strongest dungeons combine them.

What Is Random Dungeon Generation?

Random dungeon generation uses procedures to determine some or all of a dungeon's contents.

Those procedures might determine:

• Room shapes
• Corridor directions
• Number of exits
• Doors
• Stairs
• Room contents
• Monsters
• Traps
• Treasure
• Environmental features
• Themes
• Complications

The Dungeon Master may roll dice on tables, draw cards, use geomorphs, or rely on software.

Random generation has a long history in D&D.

The 1979 AD&D Dungeon Masters Guide included extensive random dungeon-generation material.

The 2014 Dungeon Master's Guide later devoted Appendix A to Random Dungeons, with procedures for generating:

• Starting areas
• Passages
• Doors
• Chambers
• Stairs
• Connections
• Room contents

The current 2024/5.5e Dungeon Master's Guide does not retain that Random Dungeons appendix.

Its Appendix A is a Lore Glossary, while Appendix B provides maps.

That does not mean random dungeon generation disappeared from the hobby.

It simply means current D&D does not provide the same built-in procedural appendix.

Dungeon Masters who want procedural generation today can use:

• Older D&D material
• Third-party tables
• Random generators
• Geomorphs
• Their own custom procedures

What Is Handcrafted Dungeon Design?

Handcrafted design means the Dungeon Master intentionally creates the dungeon's structure and contents.

That may include deciding:

• Why the location exists
• Who built it
• How spaces connect
• Who occupies it now
• Which factions interact
• Where important clues appear
• Which routes provide shortcuts
• Where major hazards exist
• How the environment changes
• What information players can discover

"Handcrafted" does not necessarily mean drawing every five-foot square personally.

A Dungeon Master might still use:

• Published maps
• Dungeon-mapping software
• Premade encounters
• Random names
• Monster generators
• Treasure tables

The defining feature is that the final arrangement is intentionally selected for the dungeon being built.

Random Does Not Mean Unplanned

One of the biggest misconceptions about random generation is that the Dungeon Master must accept every result exactly as rolled.

You do not.

A random table can serve as:

• Inspiration
• Constraint
• Prompt
• Starting point
• Tie-breaker
• Surprise generator

Suppose your table produces:

Temple

Underground river

Prison cells

Giant spiders

Abandoned workshop

You do not need to place those five ideas into unrelated rooms and call the dungeon complete.

Instead, ask:

Why are these things together?

Perhaps an old religious order imprisoned enemies beneath its temple.

An underground river once powered machinery in the workshop.

After the complex was abandoned, giant spiders entered through the river caverns and now occupy the prison level.

Random results created the ingredients.

The Dungeon Master created the relationships.

That is still random-assisted design.

Handcrafted Does Not Mean Predetermined

The opposite misconception is that deliberate dungeon design means scripting what players will do.

It does not.

You can intentionally design:

• Multiple entrances
• Loops
• Secret routes
• Factions
• Shortcuts
• Optional areas

without deciding how players must navigate them.

For example, you might deliberately place:

• A guarded front entrance
• A collapsed wall accessible by climbing
• A drainage tunnel
• A secret route revealed by a prisoner

You designed every one of those paths.

The players still decide which one matters.

Handcrafted geography can support strong player agency when it creates meaningful alternatives.

The Real Difference: Where Does the Surprise Come From?

Random generation introduces surprise during preparation or even during play.

Handcrafted design concentrates surprise on the players while the Dungeon Master knows more of the structure in advance.

Neither arrangement is inherently superior.

Some Dungeon Masters enjoy discovering a location as they generate it.

Others enjoy knowing the dungeon deeply enough to portray its relationships consistently.

Some use both.

A DM might randomly generate the old crypt's layout, then spend thirty minutes deciding:

• Who built it
• Why one corridor collapsed
• What faction moved in
• Where the clues belong
• How the location connects to the campaign

The geometry remains surprising.

The final dungeon becomes intentional.

Where Random Generation Is Strongest

Breaking the Blank Page

Sometimes the hardest design problem is simply starting.

A blank sheet asks:

What should this dungeon be?

A table might instead give you:

Flooded shrine.

Now you can ask:

• Why is it flooded?
• Who worshiped here?
• What survived underwater?
• Is the flooding recent?
• What lies below the water?

The random result does not finish the design.

It gives your imagination something to react to.

Producing Ideas You Would Not Choose Yourself

Dungeon Masters develop habits.

You might repeatedly create:

• Ruined temples
• Goblin lairs
• Abandoned mines
• Straight corridors

Random procedures can interrupt those patterns.

Perhaps you roll:

• Greenhouse
• Vertical shaft
• Observatory
• Fungus colony
• Political prisoner

That combination may lead somewhere you would never have reached from a blank page.

Creative constraints can be useful precisely because they remove some choices.

Procedural Exploration

Some campaigns intentionally make procedural uncertainty part of play.

Examples include:

• Shifting magical dungeons
• Endless labyrinths
• Unexplored underworlds
• Procedurally generated ruins
• Solo exploration
• Roguelike-inspired adventures

In those situations, randomness is not merely a preparation shortcut.

It is part of the dungeon's identity.

Players may know that the environment itself is unstable or unknown.

Improvisational Support

Players sometimes enter an area you have not detailed.

A random table can quickly answer:

• What is in this chamber?
• Who lives here?
• What strange feature exists?
• What complication appears?

That is particularly useful when you need raw material immediately.

But random generation should still respect already established facts.

If the party is exploring an underwater ruin, a result such as:

Dry wooden library full of burning braziers

may need adaptation.

Perhaps it becomes:

A sealed archive whose magical wards kept the interior dry.

Or perhaps you simply reroll.

Randomness does not override continuity.

Replayability and Variation

Procedural tools can produce many different maps and combinations from the same underlying system.

Donjon's current 5.5 Random Dungeon Generator, for example, allows Dungeon Masters to adjust factors such as:

• Dungeon level
• Party size
• Motif
• Size
• Layout
• Room size
• Doors
• Corridors
• Stairs
• Map style

That makes procedural generation useful when you need several distinct locations built from a common workflow.

The results are different.

Whether they are interesting still depends on how they are interpreted and used.

Where Random Generation Needs Help

Architectural Logic

Random geometry does not automatically understand why a building exists.

A fortress usually needs:

• Defensible entrances
• Routes for inhabitants
• Living areas
• Storage
• Water
• Sanitation
• Defensive positions

A monastery may need:

• Worship spaces
• Sleeping quarters
• Food preparation
• Storage
• Records

A random layout may accidentally produce something plausible.

Or it may not.

If architectural logic matters, curate the result.

Ask:

Could the people who used this place actually have used it this way?

Fantasy allows strange architecture.

It still benefits from internal logic.

Theme

A random result can create thematic noise.

Suppose the dungeon is an ancient frost giant burial complex.

The generator produces:

• Fire elemental
• Tropical garden
• Halfling merchant
• Clockwork workshop

Those elements could become fascinating if you develop explanations.

But if no explanation exists, the dungeon may feel assembled rather than discovered.

The solution is not rejecting randomness.

It is filtering through the dungeon's theme.

Faction Coherence

Creatures should not merely occupy rooms.

Ask:

• Who controls this area?
• Who hates whom?
• Who supplies food?
• Which routes are patrolled?
• Who avoids certain rooms?
• What happens if the party changes the balance?

A random encounter table can suggest inhabitants.

The DM can connect them into a living environment.

Information Placement

If the campaign requires important information, do not leave its existence entirely to chance.

Suppose players need to learn:

The duke's adviser financed the excavation.

You may place several clues intentionally.

Random generation can determine surrounding rooms.

But campaign-critical information should generally be resilient enough to survive:

• A skipped room
• A failed check
• An unexpected route

Randomness can support discovery.

It should not make the campaign depend on one lucky procedural result.

Where Handcrafted Design Is Strongest

Campaign-Specific Locations

Some dungeons carry major campaign meaning.

Examples:

• Villain's headquarters
• Character's ancestral home
• Lost temple central to the campaign
• Prison containing an important ally
• Vault holding a major artifact

In those cases, intentional design can help every part reinforce the campaign.

The villain's headquarters might reveal:

• Priorities
• Resources
• Allies
• Vulnerabilities
• History

The location itself becomes characterization.

Environmental Storytelling

If you know what happened in a location, you can deliberately distribute evidence.

Suppose a dwarven mine was abandoned after workers uncovered something ancient.

The upper levels might show:

• Interrupted work
• Missing equipment
• Barricades
• Warning inscriptions

Deeper areas might reveal:

• Emergency defenses
• Collapsed passages
• Failed containment measures

The final chambers explain what they were trying to contain.

Players reconstruct the history through exploration.

That type of information progression benefits from intention.

Meaningful Spatial Relationships

A handcrafted dungeon lets you deliberately design:

• Loops
• Shortcuts
• Sight lines
• Vertical connections
• Alternate entrances
• Secret passages

Those relationships can create strategic mastery.

Players gradually learn:

"The chapel stair comes out behind the barracks."

"The drainage tunnel reaches the lower level."

"If we open this gate, retreat becomes much easier."

That kind of geography can be generated randomly too.

Handcrafting simply gives you more direct control over it.

Thematic Set Pieces

Sometimes you have a specific image in mind.

Perhaps:

A ruined throne room split by a chasm, with the throne still standing on an isolated pillar of stone.

Or:

A library whose shelves slowly rotate as a magical mechanism rearranges the room.

Those rooms benefit from deliberate construction because the location itself is the idea.

Use random generation around them if you want.

The centerpiece can remain handcrafted.

Where Handcrafted Design Can Go Wrong

Overinvestment

The more hours you spend on one location, the harder it may become emotionally to let players bypass it.

That can lead to:

• Moving encounters into the party's path
• Blocking unexpected solutions
• Preventing shortcuts
• Forcing players toward prepared material

Unused preparation is not failure.

Players are allowed to miss rooms.

If avoiding a chamber required clever play, skipping it may actually demonstrate that the dungeon worked.

Designing a Route Instead of a Place

A handcrafted dungeon can become:

Room A → Room B → Room C → boss.

This may be intentional.

A short gauntlet can work.

But if you want exploration, consider:

• Loops
• Branches
• Alternate entrances
• Shortcuts
• Faction routes
• Secret passages

Do not mistake intentional design for linearity.

Perfectionism

Handcrafting can encourage endless polishing.

You may start asking:

• Is this corridor realistic enough?
• Is this map beautiful enough?
• Does every room have a brilliant gimmick?
• Do I need another clue?
• Should I redraw the entire lower level?

At some point, the dungeon needs to be playable rather than perfect.

The players will change it anyway.

The False Comparison: Which Is Faster?

There is no reliable universal time ratio.

A random layout might take seconds to generate.

Then you may spend hours making it coherent.

A Dungeon Master experienced with sketching may handcraft a six-room dungeon quickly.

Another may spend several evenings creating a polished VTT map.

Preparation time depends on:

• Size
• Detail
• Mapping method
• System
• Experience
• Encounter preparation
• Visual assets
• How much generated content requires curation

Random generation can reduce blank-page work.

It does not guarantee lower total preparation.

The False Comparison: Which Produces Better Stories?

Neither.

Story emerges from what happens during play.

A randomly generated chamber can become unforgettable because the players:

• Befriend its inhabitant
• Trigger a catastrophic trap
• Discover an unexpected connection
• Make a difficult decision there

A meticulously designed throne room may be forgotten if nothing meaningful happens.

Design influences opportunities.

Players and consequences create memories.

The False Comparison: Which Supports More Agency?

Neither method automatically does.

A random dungeon can be:

• Linear
• Branching
• Interactive
• Restrictive

A handcrafted dungeon can be the same.

Agency comes from:

• Meaningful choices
• Useful information
• Consequences
• Alternate approaches
• Responsive situations

The source of the map does not decide that.

The Strongest Option Is Often Hybrid Design

Random generation and handcrafted design work especially well together.

Here are several practical hybrid approaches.

Method 1: Generate the Skeleton, Design the Meaning

Use a random generator for:

• Overall layout
• Room shapes
• Corridors
• Stairs

Then intentionally determine:

• Function
• Inhabitants
• Factions
• Clues
• Hazards
• Shortcuts
• History

This is useful when map geometry slows you down but worldbuilding does not.

Method 2: Design Anchors, Generate the Gaps

Handcraft several important areas.

For example:

1. Entrance
2. Broken excavation chamber
3. Faction headquarters
4. Ancient seal
5. Final vault

Then generate or improvise secondary spaces around them.

The important campaign material remains intentional.

The surrounding dungeon gains unpredictability.

Method 3: Random Stocking, Handcrafted Map

Draw a dungeon layout you like.

Then use tables to determine some room contents.

Perhaps you randomly determine:

• Empty
• Creature
• Hazard
• Treasure
• Strange feature
• Evidence

Then adapt each result to the location.

This gives you control over geography while introducing content you did not personally select.

Method 4: Random Prompt, Handcrafted Execution

Roll only for concepts.

For example:

Shrine + imprisoned creature + flooding

Then design the entire room intentionally around that prompt.

Randomness provides inspiration.

Craft determines execution.

Method 5: Generate, Then Curate

Generate a complete dungeon.

Then review it.

Ask:

• Which rooms are redundant?
• Which connections are boring?
• Where does a loop help?
• Which creatures do not fit?
• Where is information missing?
• Which area deserves a stronger identity?

Delete.

Combine.

Move.

Rewrite.

The generated map becomes a draft.

Method 6: Build Custom Tables

Generic tables produce generic possibilities.

Campaign-specific tables can be much stronger.

Suppose the dungeon is a drowned monastery.

Instead of rolling:

Random monster

create a table containing:

• Drowned acolytes
• Sahuagin scouts
• Escaped guardian
• Rival salvage crew
• Giant eel
• Frightened pilgrim spirit

Now randomness remains consistent with the location.

You have built theme into the procedure.

Method 7: Use Randomness Only When You Do Not Care Which Answer Occurs

This may be the most useful rule.

Ask:

Does this decision need a particular answer?

If yes, choose intentionally.

If no, randomness may be ideal.

Examples:

Choose intentionally

• Location of the campaign-critical clue
• Identity of the faction leader
• Cause of the dungeon's collapse
• Route connecting two major areas

Randomize if useful

• Minor room dressing
• Which patrol appears
• Decorative treasure
• Name of an incidental NPC
• Weather
• Which unused chamber contains a supply cache

Randomness is strongest when several outcomes would all be acceptable.

Example: The Collapsed Dwarven Mine

Suppose the party discovers an abandoned dwarven mine.

You know three things must be true:

1. The dwarves accidentally breached an ancient structure.
2. The mine was evacuated suddenly.
3. Something still lives below.

Those are intentional.

Random Step 1: Generate Layout

A generator gives you:

• Central shaft
• Two looping mine routes
• Several dead ends
• Lower cavern level
• Secondary exit

You keep most of it.

Handcrafted Step 2: Establish History

You decide:

• Upper tunnels are normal mining areas.
• Lower excavations show increasingly strange stonework.
• One breach connects to architecture far older than the mine.

Random Step 3: Generate Prompts

You roll several contents:

• Abandoned equipment
• Monster
• Water
• Shrine
• Treasure

Handcrafted Step 4: Interpret

Abandoned equipment

Emergency excavation gear discarded during evacuation.

Monster

Blind predators moved into the abandoned shafts.

Water

A broken underground channel flooded one route.

Shrine

A dwarven memorial built after the first miners died.

Treasure

Emergency payroll never recovered during evacuation.

The dungeon is both generated and designed.

The random results gave you ideas.

Intentional interpretation made them belong together.

Random Generation During Play

Generating the dungeon while players explore creates a different experience from generating it before the session.

It can work particularly well for:

• Solo play
• Magical labyrinths
• Infinite dungeons
• Procedural campaigns
• Improvisational exploration

But remember an important limitation:

Once something becomes true at the table, it becomes part of the world.

If the generated northern passage ends at a chasm, remember that chasm.

If the party discovers goblins on Level 2, consider what those goblins need and where they travel.

Procedural generation creates facts.

Campaign continuity should preserve them.

Otherwise randomness feels arbitrary rather than exploratory.

Current Random Dungeon Tools

Donjon

Donjon remains one of the best-known free collections of random tabletop tools.

Its current 5.5 Random Dungeon Generator can generate and export dungeon material with options for:

• Dungeon level
• Party size
• Motif
• Dungeon size
• Layout
• Room arrangement
• Doors
• Corridors
• Stairs
• Grid
• Map style

It can currently export formats including:

• Images
• Printable maps
• HTML
• PDF
• JSON

Donjon is useful for:

• Fast drafts
• Procedural campaigns
• Inspiration
• Layout generation
• Disposable side locations

As with any generator, review the output before treating it as campaign canon.

Older D&D Random-Dungeon Procedures

If you want to examine traditional D&D procedural dungeon generation, two historically important sources are:

1979 AD&D Dungeon Masters Guide

Gary Gygax's first-edition Dungeon Masters Guide included extensive random dungeon-generation procedures.

2014 Dungeon Master's Guide

Appendix A, Random Dungeons, provides fifth-edition procedures for generating maps and contents.

2024 Dungeon Master's Guide

The 2024/5.5e DMG changed direction.

It does not include the 2014 Random Dungeons appendix.

Appendix A is now the Lore Glossary.

Appendix B contains dungeon and structure maps that Dungeon Masters can use or adapt.

If you want procedural generation alongside the current rules, use an external generator, older procedural material, or your own tables.

Handcrafted Design Tools

Handcrafting does not require pencil and graph paper, although that remains perfectly viable.

You can use:

• Dungeon Scrawl
• Dungeondraft
• VTT map tools
• Published maps
• Geomorphs
• Hand sketches
• Image editors

The tool does not determine whether the dungeon is handcrafted.

The intentional relationships between its parts do.

Choosing the Method for a Campaign-Critical Dungeon

For a location carrying major campaign information, I usually recommend more intentional curation.

That does not mean every corridor must be handcrafted.

Instead, identify what cannot safely be left to chance.

For example:

Must Be Intentional

• Why the place exists
• Main factions
• Major revelations
• Important relationships
• Essential information
• Consequences

Can Be Randomized

• Some room shapes
• Minor contents
• Incidental treasure
• Wandering creatures
• Decorative details

You preserve the dungeon's purpose while still gaining surprise.

Choosing the Method for an Optional Side Location

A side ruin might tolerate much more procedural generation.

Perhaps the party decides unexpectedly:

"We want to explore those standing stones."

You had not prepared them.

A random generator can give you:

• Layout
• Several rooms
• Creatures
• Treasure
• Strange features

Then spend a few minutes asking:

Why does this place make sense here?

That may be all the preparation needed.

Choosing the Method for a Megadungeon

Large dungeons are excellent candidates for hybrid techniques.

Handcraft:

• Major zones
• Factions
• Entrances
• Important landmarks
• Major vertical connections

Generate or procedurally stock:

• Minor rooms
• Wandering encounters
• Dungeon dressing
• Some treasure
• Lesser complications

The goal is sustainability.

You do not need to personally invent every barrel in a 200-room complex.

Choosing the Method for Improvisation

When players move beyond your preparation, use randomness to answer questions you do not already care about.

Suppose they enter an abandoned district beneath the city.

You know:

• Smugglers once used it.
• Something drove them out.
• A tunnel connects to the river.

You do not know what every chamber contains.

Roll or generate as needed.

Then interpret each result through those three truths.

That prevents randomness from feeling disconnected.

Match the Method to Your Own Brain

Some Dungeon Masters enjoy blank-page design.

Others hate it.

Some enjoy rolling on tables and connecting strange results.

Others find that additional mental work exhausting.

Some love detailed maps.

Others want three bullet points and a sketch.

There is no virtue in choosing the more labor-intensive method.

A sustainable preparation method is better than an impressive one you cannot maintain.

Use the process that helps you arrive at the table prepared enough to respond to the players.

Do Not Assume Player Preference From the Method

Players generally experience the dungeon, not your preparation technique.

They may never know whether a room was:

• Generated randomly
• Handcrafted
• Adapted from a module
• Inspired by another map
• Improvised

What matters to them is whether they can:

• Understand the situation
• Make meaningful choices
• Discover interesting things
• Affect what happens

A random dungeon can provide that.

So can a handcrafted one.

How The Dungeon Trainer Fits Into This Workflow

The Dungeon Trainer can support both approaches, but it does not currently replace a dedicated tactical dungeon generator or mapping application.

Generators

The platform can generate:

• Locations
• Monsters
• Characters
• NPCs
• Magic items
• Spells

These provide useful raw material for procedural or hybrid design.

For example, you might:

1. Generate a location concept.
2. Generate several inhabitants.
3. Choose which pieces belong together.
4. Place them on a map you created elsewhere.

The platform generates ingredients.

The Dungeon Master establishes the dungeon's spatial structure and relationships.

Forge

Forge can build a campaign or one-shot from a premise, themes, and tone.

That can help establish:

• Campaign context
• Adventure direction
• Supporting material

It is not the same thing as generating a coherent room-by-room tactical dungeon floor plan.

Use dedicated mapping tools when spatial dungeon architecture is the actual task.

Scry

Scry can help after either random or handcrafted preparation.

Suppose you have created a dungeon entrance with:

• Guarded gate
• Drainage tunnel
• Cliff access

Your notes assume the party enters through the gate.

Scry can help expose plausible alternatives and preparation gaps.

That matters regardless of how the map was created.

The Coach

The Coach can help practice skills relevant to either design style:

• Exploration adjudication
• Player agency
• Pacing
• Information clarity
• Unexpected solutions

Random generation often demands improvisation.

Handcrafted design often requires letting go of expected outcomes.

Both benefit from those skills.

Campaign Memory and Codex

Procedural campaigns particularly benefit from reliable continuity.

Once a randomly generated:

• NPC
• Location
• Faction
• Quest
• Event

enters play, it becomes part of campaign history.

The Dungeon Trainer maintains searchable campaign records while keeping unrevealed Dungeon Master information hidden.

That helps random discoveries become persistent world elements rather than disposable results.

A Practical Decision Framework

When deciding what to randomize, ask:

Does this fact need to support a specific campaign purpose?

If yes, design or curate it intentionally.

Would several outcomes be equally acceptable?

If yes, randomness may save time or create inspiration.

Will players make strategic decisions based on this feature?

Make sure the result remains internally consistent once established.

Is this campaign-critical information?

Do not rely on one random result to make it available.

Am I stuck because the page is blank?

Randomize something.

Am I accepting a random result I know makes no sense?

Change it.

You are the Dungeon Master.

The table is a tool, not an authority.

The Best Workflow Is Often Uneven

You do not need to use the same ratio for every dungeon.

One location might be:

• 90% handcrafted
• 10% random

Another:

• Random map
• Handcrafted factions
• Random room dressing
• Intentional clues

Another may be generated almost entirely during play.

That flexibility is the real advantage of learning both approaches.

You stop asking:

"Am I a random-generation DM or a handcrafted-design DM?"

and start asking:

"Which parts of this dungeon deserve my attention?"

That is a much more useful preparation question.

Conclusion

Random dungeon generation and handcrafted dungeon design solve different problems.

Random generation is excellent for:

• Breaking creative blocks
• Producing unexpected combinations
• Supporting procedural play
• Providing improvisational material
• Generating layouts quickly

Handcrafted design is excellent when:

• Spatial relationships matter
• Campaign-specific information must be placed intentionally
• Environmental storytelling matters
• Factions need deliberate relationships
• A particular set piece or theme deserves focused design

Neither method guarantees:

• Better pacing
• Better agency
• Better stories
• More memorable encounters

Those depend on what you do with the results.

The strongest workflow is often hybrid.

Generate what you do not need to control.

Design what needs intention.

Curate everything that becomes part of the campaign.

And once the players enter the dungeon, stop worrying about which parts came from dice and which parts came from you.

At that point, the dungeon belongs to the table.

Sources and Further Reading

Alexander, J. (2010). Xandering the Dungeon. The Alexandrian. https://thealexandrian.net/wordpress/13085/roleplaying-games/xandering-the-dungeon

Donjon. (2026). 5.5 Random Dungeon Generator. https://donjon.bin.sh/5e5/dungeon/

Gygax, G. (1979). Dungeon Masters Guide. TSR.

Shea, M. E. (2018). Return of the Lazy Dungeon Master. Sly Flourish.

Wizards of the Coast. (2014). Dungeon Master's Guide. Wizards of the Coast.

Wizards of the Coast. (2024). Dungeon Master's Guide. Wizards of the Coast.

The Dungeon Trainer. (2026). What Is a Dungeon Crawl in D&D and How Does It Work? https://thedungeontrainer.com/blog/what-is-a-dungeon-crawl-in-d-d-and-how-does-it-work