I'm a new GM and I need to undestand how to build enjoyable locations.

I don't get how to build interesting dungeons. For example, I need to explore a castle. But a castle is not as complicated as the classical "maze of caverns", so it would just be a boring walk. What can I do to improve my dungeon creating skills?

  • 2
    \$\begingroup\$ Edited. Is it ok to post two questions in a small period of time? \$\endgroup\$
    – olinarr
    Commented Feb 18, 2014 at 15:49
  • \$\begingroup\$ Generally, yes. \$\endgroup\$ Commented Feb 18, 2014 at 15:51
  • 5
    \$\begingroup\$ Not even just generally, it's encouraged! There is such thing as too many questions posted at once, but the limit is up around ten, not two, so not worries. (And the limit is more because, you can't expect anyone, including yourself, to pay quality attention to that many questions all at once.) \$\endgroup\$ Commented Feb 18, 2014 at 16:27
  • 5
    \$\begingroup\$ Just noting, this topic is probably a little too broad - entire books have been written on dungeon creation. Consider focusing on some specific subset of that. Though, answerers, I will note he asked for "how to improve his skills" which might be better answered with references like that instead of your own brainstormed lists of "three things that make a good dungeon." \$\endgroup\$
    – mxyzplk
    Commented Feb 18, 2014 at 18:23

6 Answers 6


For me, there are 2 main questions that you should ask yourself. The first one is "what will I, the GM, find interesting?" The second is "what will interest my players?" These are the most important questions and from them we are going to build the dungeon.

After we do that, there's one thing that should be made clear: "Where the main action in this dungeon should take place? Is it a long road filled with monsters or is it a collection of rooms, each one with its own set of monsters (or even a single one)?" If it is a collection of rooms with a few corridors connecting between them, you can practically almost skip the walking between the rooms, focusing instead on the rooms themselves. As the question is about a castle, I'll assume that we're talking about a set of rooms with monsters and the like within them, connected by corridors.

Our next question, then, is "what is the story of this castle?" The castle is not just a castle, right? It's a famous place, the source of many a legend. As such, we need to create such a legend. Let's go for a simple one, as an example: "The famous wizard Olivius created this dungeon in order to defend all of his treasures, but most of all, in order to lock the great dragon Zarius Maximus away from the village, believing he may be able to catch 2 birds at once." What does this legend tell us? That there's a dragon waiting in the end of the dungeon, that this place was made by a great wizard and as such will be guarded by many dangerous things, and that there's a great treasure there. It also gives us a reason to go nuts with what we're about to put there.

Now we implement the answers to the first 2 questions. If, for example, Robert is interested in big fights against huge monsters, we can put a giant for him. If Lisa wants to disarm a great trap, we can implement a trap that will be so thrilling that when she'll disarm it it will be remembered for ages. The trick is to put in the rooms what the group (and that includes you) finds interesting.

When we finish this we have a list of rooms with monsters/traps/treasures within them. Now we're beginning to connect between them. The main trick here is to make sure that we don't forget anyone of the players. It means that each room should be surrounded by rooms that interest other players, aiming for an equal representation of all the players' interests. Then you tweak it a little, so it won't be too formulaic. If the rooms that interest Lisa the most are always surrounded by rooms that interest Olivia, then we'll replace some of the surrounding rooms by rooms that interest Richard.

To conclude all of this, here's a summary of the steps you should make:

  1. What does interest me? What does interest my players? These are the things that we want to see within the dungeon.

  2. Is this a long road filled with monsters or is it a collection of rooms, each one with its own set of monsters?

  3. What is the story/legend of this dungeon?"

  4. Now we fill the dungeon with monsters, traps and the like, according to the legend and to the things that interest us.

  5. Connecting the different parts of the dungeon with corridors, for example.

  6. Tweaking the way we located the rooms a little bit, so it won't be formulaic.


A dungeon is a sort of "container" for challenges. (The types of challenges you can include are combat, puzzles/traps, and negotiation. Build challenges according to your group's desires!).


All dungeons have description - details about what makes them interesting - think of walking through an abandoned building or old set of ruins - what can you simply deduce by looking at the place? What kind of people lived there? What is in ruin, how? Do you see a grand statue of intricate design, carved by patient, skilled hands, now sitting under an open sky where a trickle of water spills upon it?

You might be good at making up a lot of detail with a few words or sentences to give you inspiration written to yourself: "Old stone", "Greek-ish meets Aztec design", "Constant ruin of water"

Or you might want to have photos or write up area descriptions as many modules/adventures do.

You will want to talk to the group and let them know if your games are supposed to have a lot of hidden/secret stuff or not - because if so, or if the players THINK there's supposed to be, they'll spend hours in each room poking at descriptive stuff because they're paranoid about everything either being a trap or a hidden treasure they overlooked.


The actual map, depends on whether your group finds exploration interesting or not.

If your group doesn't, a lot of times what I will do is not draw a map for the whole complex - just a few key rooms/areas where combat might happen, and sum it up thus, "It takes a few hours of careful searching, you find yourself deep in cavernous room where the Night Shrine still stands..."

But if your group does enjoy exploration, you can do a lot to make maps more fun.

First, a lot of ancient underground building was kind of hap-hazard. Especially if a structure had parts tunneled out over generations or centuries, so the weirdness of the classic dungeon isn't entirely without any basis.

Second, if you're going to use an area that normally would have a boring, straightforward layout (such as symmetrical layout, etc.) consider what ruin might disrupt that symmetry and break up the ability to move from one area to another. Roof collapses, cave ins, giant fissures in the ground, flooding, giant tree roots, etc. Take your "pre-ruin" map and then draw lines indicating places you can't cross, or crossing would be it's own challenge ("...sure, you could get over it. It's gonna take about 20 minutes and a bit of climbing. Just remember, it's going to probably be just as hard to get back and you're not sure WHAT is on the other side...")

Layout Tricks

There's some features that make dungeon layouts fun (if you like to take videogames for inspiration, "Metroidvania" style game maps or Mario/Zelda games use a lot of these tricks):


Being able to see a place you want to get to, but you can't, just yet. This might be a treasure chest on the other side of chasm. A door at the end of a hall, that's blocked off by a locked gate. Previews are fun because they give the players a sense of things to look forward to and as they go along, they're going to be trying to figure out how to get there and piecing together the map.

Gated Zones

Tied into previews, there may be places you can't get to because you need something first - a key, a rope, finding a way to unlock a gate, a means of walking through the hallway that's perpetually on fire, etc. Gated Zones gives players a reason to come back to an area ("Wait, we could use this to get past the river we saw in the east section..."). It rewards players who like to keep things in mind.

A fun type of gate zone is to have parts of the dungeon which are "unlocked" by permanently changing them - pushing down part of a wall, knocking down a support pillar to serve as a bridge, etc. It lets the players "put their mark" on the dungeon, and also changes the tactical setup of the place... "Crap, now the zombies can chase us through that...").

Be warned that obstacles that matter in the early levels tend to get bypassed with ease as characters get more power - flying, teleporting, passing through stone type abilities start to appear and dungeons stop serving as challenges.


It's always fun to find a way to get somewhere that's shorter, safer, than the long way. It also helps players start mentally connecting areas they thought were separate and getting a better feel for the layout of the place.

A short cut might be something like a door they unlock or a secret door, or it might be something like finding a wall that crumbles apart allowing them to crawl through to another spot.

Short cuts are most rewarding when players find a shortcut BACK to a place they've been before. The more pain you had to go through getting there in the first place, the sweeter the reward is to having a quick way back without having to expose yourself to danger again.

One Way Passages

One way trips are a great way to make a dungeon, a dungeon. This could be the obvious sorts of contrived traps like slides, pitfalls, or doors that lock behind you, but I generally like more naturalistic ways of making one-way passages - cave ins, weak floors, crumbling bridges, etc. It also emphasizes the general ruin and scariness of the place.

You do need to be careful about setting these up in ways that split the party. Not only will this make encounters much more difficult to balance, it also means that when action is happening for one part of the group, the other group is sitting around doing nothing.

Mysteries and Reveals

This is the best stuff, but also the hardest to pull off. You set up clues about something throughout the dungeon, and when the players get to a certain point, the reason of something is revealed.

This is not "And NPC X says, 'This is why it happened'", no, this is stuff like finding huge roots coiling through the entire dungeon, having to crawl over them, under them, deal with monsters nested around them, finding entire passages crushed or broken by them, and then, deep in the dungeon finding a magical laboratory with a few "Plant Growth" scrolls and a magic circle and several partially crushed skeletons under the roots and realizing it all was magical experimentation gone horribly wrong.

But this can be small things too! "Why can't we open the door?" later, get to the other side, "Oh, geez, a statue toppled in front of it."

Again, though, this requires players who are interested in exploration, otherwise you end up prepping all this stuff and all the players remember is "Oh, there was an L-shaped room where we fought the big spider". So talk to your players before putting in work.

  • 3
    \$\begingroup\$ Tied into a few of the points in this answer is the concept of retrocursive location design, which is to say, giving the players reasons to come back to old areas later and ways of making the locations still interesting when they do. \$\endgroup\$
    – GMJoe
    Commented Feb 19, 2014 at 4:12
  • \$\begingroup\$ Great ideas! Another dungeon challenge that adds tension to the game is the "cat and mouse" scenario, where the PCs are being pursued by dungeon denizens. This could be a nasty, big minotaur, or a squadron of Orc guards on alert after the PCs set off an alarm of some kind. Or even a slow but relentless pursuit by gelatinous cubes that sweep through the dungeon hunting for PC parties like the ghosts in Pac-man. \$\endgroup\$
    – RobertF
    Commented Jun 16, 2014 at 20:45
  • 3
    \$\begingroup\$ @RobertF That only works if your PCs are smart enough to run away when they're supposed to. ;) \$\endgroup\$ Commented Jul 2, 2014 at 20:02

I've been a GM for a long time, and a few things I've learned are that dungeons should feel authentic, be hand-crafted with a purpose in mind, and ferry between encounters.

Make Dungeons Authentic

Dungeons aren't usually built by bad guys for bad guys to lurk in and then raid the surroundings from, and even if they are you've got some things to consider. Caves, for instance, will likely be expanded by hand or with tools/machinery (based on tech and capabilities of the inhabitants) in order to have more storage, some security measures, and dwelling spaces. Normal caves aren't all that great, since they tend to be somewhat wet and somewhat hazardous to maneuver in, so they'll have pathways and such worn or intentionally cut through them, and anything in them will be protected from the humidity somehow, unless it's fine just sitting there (i.e. no iron stuff sitting in plain sight unless it's rusting over).

I like to use four basic requirements for my dungeons (which also applies nicely to pretty much every other structure for adventures): functionality, storage, needs, security.


Every building has a purpose. Even if it's a monument, it's meant for something. Who built the dungeon? Why? Does it do something that the current inhabitants might not like, and have they tried to suppress this?

For instance, if goblins move into a human castle, the will likely deface the monuments, because they really don't care about the former human inhabitants. Most dungeons have a function ostensibly of being a defensive or offensive position, so anything not on a hill will be concealed, and anything on a hill will be fortified. Quarries, on the other hand, may be lightly fortified, especially if the ores being mined are valuable, or they're in a dangerous region, but are unlikely to be in a strategically advantageous position.

A good example for this would be a medieval church; they served obviously as a symbol of faith and a meeting place for the locals, so they're made of long-lasting stone and built tall so they can be seen. Many also have bell towers of various sorts, which can be used for communication of important events, holidays, or emergencies. You have a sanctuary for meeting, various clerical rooms (depending on time, local customs, and whether the priest lived in the church or off-site), and a spire in pretty much every church worth its salt.


I know of absolutely no business that does not keep its supplies on site. No military base has to go out to grab ammunition on a daily basis. The church in my example above likely has storage for candles, clergyman's garments, and various materials used in worship and spiritual services.

Any major (and even most minor) sites prior to mass transit will have storage for food and water. In the modern day, water's been offset by plumbing, so we usually don't store it except in central places, but we usually can find some food stored pretty much everywhere, even if it's just in a worker's cubicle for a mid-day snack.

Also consider weapons and the like. A creature's lair will likely have little if any storage, even the least sophisticated humanoids will have some storage for necessities, and hoarders (in mind or culture) like dragons or dwarves will build massive vaults, burrows, and stockpiles for goods.


Every once in a while, when you're building a place, a pressing need comes up. Medieval castles were often built of wood, not because it was superior but because it was there. If there are no quarries around you can bet that large structures will be made from brick rather than stone. Rough terrain may even mean that fortifications for a provincial region fall outside the main economic areas, since there's no good way to build effective fortifications with the resources available.

Note that when I say needs I'm referring to situational needs, rather than inhabitants' needs. Look at modern buildings that are often built to be earthquake/wind resistant. Even dungeons in caves will have supports and the like because they're supposed to stay standing rather than burying their inhabitants.


Finally, security is a major part of many dungeons. Consider things that the inhabitants would find reasonable. It's entirely possible that the door to the treasure vault is trapped, but you won't find a trap at the front gate or whatever the inhabitants use most, because it has a way too likely chance of accidentally killing the users.

Security rather typically focuses on denying entry. Rope bridges across chasms work just as well as a gate against many threats, so long as you aren't accidentally trapping yourself in view of archers or mages. Heavy doors are preferred to a flimsy door with a trap; adventurers are actually one of the least of the worries of many people.

Likewise, any fortified location will likely have some sort of concern for escape as well as denying entry. You want a bolt-hole for when the bad guys come, and the bad guys view the good guys as bad guys. Even if it's as simple as a safe room to hide in until you make a stealthy (or bloody) egress, dungeons likely have somewhere to hide.

Security measure that actively harm are rare if you want a "realistic" feeling dungeon, though alarms and the like may be relatively common should players do something they aren't supposed to.

Dungeons With A Purpose

More so than the architectural design of the dungeon, each dungeon should have a purpose for the players. No dungeon in a hack-and-slash campaign should exist simply to be explored; although this is certainly a viable option for exploration oriented campaigns, you just don't deliver the purpose. I like to think of each dungeon as filling separate categories of narrative and game fulfillment.

Narrative Fulfillment

Each dungeon should have some pop to it, some zest. 13th Age does this very well, with the idea that each dungeon not only had a purpose behind its creation but also has a purpose for being placed into the game.

I like each dungeon to hold either a secret, a clue, or an adversary. Sometimes I mix and match these, but especially with the more hidden ones I try to keep each dungeon relatively clear and concise. I don't make a dungeon longer than it has to be, making sure that every element in my dungeon leads to the conclusion, rather than simply having a number of individual sections that may or may not be important and are just loosely collated. Believe it or not, this is actually something that's easy to forget to do; I've had players look at me like I'm crazy after some sessions.

The important thing is that after completing a dungeon, you're immediately on another story point. You may then go to the inn and recuperate, but there's no "well, what do we do next?" going on; using dungeons with a point keeps your games "high-impact" and won't let them slow down to the point where they can dwindle and die.

Game Fulfillment

Game fulfillment comes in two forms: reward and achievement. You always want to remember that players are not merely rewarded for playing with a new +3 Broadsword, but rather that many actually enjoy being able to actively participate and show off their abilities. I include skill rolls often, especially ones where a specific character is involved. This has the added upside of forcing my more chaotic evil parties to cooperate a little, since eventually my players realize that they can't just vengeance-shank each other as soon as nobody's looking.

I try to stagger rewards through dungeons. I like my players to see all my prepared content, so I reward them a little for each room, giving awards custom tuned to each character throughout the course of the dungeon, but only one or two at a time so that there's no sudden glut. Everyone gets a little spotlight. I also like to do session spotlighting, so big things may go to a specific character now and again at the "end" of a dungeon to give a player who either did spectacularly well or whose randomly determined turn it is a little boost.

Ferry Between Encounters

Dungeons have a flow and a purpose in swords and sorcery. Where there's one adventuring party there's another, and I like to use dungeons as little adventurer meeting grounds, graveyards, and prisons where I can drop NPC's in to guide the plot.

What this means for short-term design is that I try to always include little hints and clues about what awaits in the next room, or if that's not appropriate I at least give the players a reason to move into the next room through the plot. For instance, if there's a missing person the players will typically keep going until they find them.

For long-term design, however, I make sure that there's a little something in each dungeon that indicates where the players can go next. I don't force the issue, but I'll make sure to mention that the wonderful +3 Rapier the Ranger found has a partner, forged by master duelists, and that the other is somewhere accessible to them. Of course, this isn't for everyone, but since I don't like to force my players into my plot lines I drop lots of little hints and teases into dungeons to give them a long-term goal for future adventuring.

This sounds a lot like it's a plot point, but it's not. My groups have a tendency to kill, on accident or on purpose, many NPC's, and I like to let them do so. Having another little lead for them means that I can gather up the plot again before they run out of things to do and get bored with the game, even if the next session is little more than a road-trip followed by a brawl.

The Run-Down

Don't make Skyrim dungeons. Make dungeons that feel living and vibrant, and make sense in the world. Populate each dungeon so that there are encounters that advance the story and plot, and have a significant role for the narrative. Give enough reasons for the players to keep interest in any particular dungeon, and also give them a little glimpse of a future dungeon or adventure to keep them hungry for more.

  • 1
    \$\begingroup\$ Much of this advice is custom tailored for the GM from "Be Awesome at Freelance Game Design", which I recommend for those who write their own adventures ahead of time: $5 At DriveThruRPG \$\endgroup\$ Commented Feb 18, 2014 at 16:37
  • 1
    \$\begingroup\$ You could also check out Greg Stolze's free guide to running tabletop games. It doesn't touch on dungeons, but it can help you figure out how your villains think and what sorts of dungeons they'd build/move into/summon from another plane/force the players into a mental construction of. \$\endgroup\$ Commented Feb 18, 2014 at 16:41
  • \$\begingroup\$ +1 for "realism", but I don't think this cuts to the core problem... a fully-realized, well-stocked church can still be dull if that's all it is. But this is an excellent supplement to other answers. \$\endgroup\$ Commented Feb 18, 2014 at 16:56
  • \$\begingroup\$ Yeah, I'm going to massively extend this; I had to cut it short for time. \$\endgroup\$ Commented Feb 18, 2014 at 21:57

Focus on the interesting bits

"Dungeons" are an awkward subject, since many "dungeons" I've seen have poor excuses for existing or for being as complicated as they are. Published material can be your friend here. I think studying highly-rated published dungeon adventures for what makes them successful can help you in your own creative work. Looking at bad dungeon modules and figuring out why they failed (read the reviews) can help you avoid the same traps in your work.

Edit: I'm not done reading it, but AEG's "Dungeons" seems very well written and is a comprehensive guide to creating fun and interesting dungeons. And it's very cheap for a 120 page book. http://rpg.drivethrustuff.com/product/3530/Dungeons?it=1

My short answer is, stuff has to be interesting. A complicated, twisty set of caverns is still a boring walk if there's nothing in them. A castle is the same... if the PCs need to explore it, there needs to be interesting stuff to find, stuff to interact with, stuff to be surprised by. Ideally, problems to be solved... if it's an abandoned ruin, missing sections of floor to be crossed, a party split by a stair collapsing and no clear way to join back up, etc. In short, the castle itself is just the foundation for the story... you still have to have a story that unfolds as the characters explore it. Otherwise, it's like the walk from the tavern to the market... if there's nothing interesting planned, just gloss over it with, "You explore the castle and find nothing of interest," and move on to the interesting parts of the story.

  • \$\begingroup\$ The question has been split into two separate ones, probably while you were writing it. Apart from suggesting you to always check if there are any edits on the question before posting (a nice pop-up comes out when someone does that), I also suggest you to go here and split your answer into two too. \$\endgroup\$
    – Zachiel
    Commented Feb 18, 2014 at 16:20
  • \$\begingroup\$ A lot of inexperienced dungeon designers assume that adding size automatically increases how interesting a dungeon is, without realising that that empty space has to contain interesting content in order for that to be true. +1 for pointing out the importance of interactive content. \$\endgroup\$
    – GMJoe
    Commented Jun 13, 2014 at 3:58

One of the things to consider specifically about castles is their purpose of defense. The passageways tend to be narrow so that attackers have to come in single-file and can be more readily wiped out by defensive measures such as arrow slots and oil (even non-boiling oil can be dangerous if deployed in a sloped corridor, reducing footing). Contrary to modern building design, there don't tend to be multiple paths, instead routing things through chokepoints. If you're setting things up for battle, consider where these chokepoints are, where people will set up defenses and force the attackers to undergo a withering assault while advancing.

If the PCs are visiting a castle to stock up on supplies, you don't necessarily have to make it all that interesting, of course. They can take it as a given that they can find the blacksmith, the clothier, etc. If you're looking more at intrigue, consider where people will be staying, plus how the social structure will affect the PCs. They may need to talk to Lady Winstrom, but she won't see them unless they make an appointment with her lady-in-waiting, or the PCs might arrange to run into her during her errands. Also, always fun, secret and concealed doors. While not that horribly prevalent in real castles, due to the issues of potentially letting attackers bypass your defenses, they're common in fiction, and even medieval castles in our world often had them either for servants (keeps them out of the way of the nobles in the hallways. Remember how narrow many of those hallways were) or for the defenders to move around (think passageways with arrow slits and murder holes to pour the burning pitch through). Try to make it less a matter of "Make a Notice check. OK, make a Disable Device check" but instead have an idea as to how the doors are concealed and how they might be opened.

One neat little tidbit to add in, stairs in castles should ascend in a clockwise fashion. Why? Most people are right-handed, which means that defenders at the top of the stairs will be able to freely use their right hand to attack while the attackers will be forced to either use an offhand or have more awkward angles for their blows (this can be a simple penalty to attack or bonus to defense, counterable through Ambidexterity). Similarly, you'll often find that castle roofs have little pitch to them, or are sometimes even flat, so that defenders can station archers on them. Since this is D&D, don't forget to account for how spells could be used in defense. Web isn't a horribly powerful spell, but what if it's cast across a corridor that everyone has to get through? Fireballs are notoriously dangerous things in enclosed spaces such as corridors. :) Use your imagination.


For me, the No 1 tip is: come up with the story first, then design the dungeon to accommodate it.

Another tip is to make relatively simple dungeons, and make more than one. A big dungeon is best as a set of small dungeons. Then each gaming session can have a beginning, a middle and and end, and can be memorable.

For example suppose your campaign features a quest to rid the area of goblin raiders and you're designing a goblin lair.

The story could be, the players get some clues to locate it from a mysterious ranger; they locate a secret entrance; creep in and overcome a goblin sentry without raising the alarm. Then they progress through a couple more encounters without much difficulty. Then they discover something that points to a previously unknown higher power behind the raiding; then they have a difficult battle they only just survive; finally they make it out alive with some treasure. When you've got the story, you can design the dungeon. So you sketch out a patch of forest with a secret entrance in one place and the entrance(s) the goblins usually use elsewhere. Around the secret entrance you sketch some landmarks the ranger told the party to search for, then the players can discover the entrance by following the clues. Behind the entrance you sketch a passage and some kind of sentry post, or store room, somewhere for the first encounter, perhaps with an occupied goblin they can sneak up on. Then you sketch some quarters or dorms for the easy encounters. Then you sketch the chieftan's room where they discover the new information. Then you sketch a dining hall for the big encounter where most of the goblins show up and fight. Then you sketch out a trophy room, treasury, armoury, or vault where they collect their loot. Finally you sketch the front sentry post and main entrance. After that, it's easy to add new bits and interesting features or rearrange things to accommodate other ideas you've had along the way. Tidy things up so as to avoid all the goblins hearing the players too early and spoiling the plot. A good story has to be believable, but it doesn't have to be too realistic or the story is always the same: as soon as the players invade, all the goblins come out at once and either fight to the death or surrender. :-)


Not the answer you're looking for? Browse other questions tagged .