Dungeons of 10+ rooms, how does anyone run them at all?

Since I started GMing I always enjoyed looking at large/many-roomed Dungeons or modules like Undermountain, but for the life of me I never understood how people ran them, in general, but even more so if you are running a more lethal/OSR/risky, XP for Gold style of game where players are more careful, deliberate and slower. My players and I like a good dungeon a lot! But our dungeons tend to be small and even with 3 to 5 rooms and one instance of combat a dungeon can take hours.

The problems I always had with dungeons of more than 5 and certainly more than 10 rooms:

  1. They could become one-note/boring. The almost always take the whole session, sometimes several. After two hours much less the 10th or 30th room it becomes very hard to come up with an engaging reason for the players to press on and want to explore another 5 rooms of spend entire session only in a dungeon. This is easily the top reason why I have rarely run dungeons of more than 5 rooms per session. Of course PCs could leave and come back, but that is often not very satisfying or inefficient time wise. Adding a bunch of tiny or especially empty (ish) rooms to pad the size is also not really fun for either.

  2. Limited in scope/options for players and GM. While dungeon delving (and related combat) are a huge parts of D&D a lot of other crucial, fun things can’t be done or not nearly as easily whenever you are down in a dungeon. For instance exploring different little natural areas or biomes, structures and small settlements of the world, meeting and interacting with more than 1 or 2 NPCs, buying and selling/trading goods, intrigue/factions, politics, going to a tavern all of that and more is impossible or at least much harder and rather improbable while you are exploring a dungeon. You are not going to mingle with crowds of people or interact with 3 or more NPCs in the average dungeon. It is just not that logical or realistic. So by spending the whole session (mine are 3.5 hours) or even several sessions in a dungeon you mostly give up all those other styles or parts of the game for that duration.

I am guessing this could be somewhat mitigated, I liked how Ultima Underworld had some factions or permanent inhabitants, but even there you spend the majority of your time just going from room to tunnel to room, alone (with your party) and not really feeling part of “society”.

To put it another way, every big city will have some small or big tunnels, sewer systems and dungeons (underneath castles or large churches) you can explore. A small or big dungeon will rarely have a city above it or very close by and even if it did, the players would still have to get all those cities before they could explore the dungeons. A lot of dungeons are “stand-alone” or rather far from cities.

So is less actually more? Are big dungeons holdovers for when D&D first started and was more of a strategy/tactical game and hence more focused on Dungeons (it is in the name) ? How do you and other GMs mitigate my two issues?


Hello again. I’ve discussed my way of running in a different topic. But basically you group rooms into sets that have a similar goal or theme. Within those sets of rooms the characters can move freely, so you don’t have to narrate “you go to this room, you open the door” every time. I merge rooms that are basically the same together into one. That way each room is at least a little unique.

For the second point, I think dungeons should feel isolated. But I always include some fun NPC. (like a gollum or snake person). So that PCs have someone to interact with and not only fight. A dungeon in my world would definitely have ecology and at least two biomes. Goblins and skeletons fighting for land, something like that. Random encounters can also help spice up dungeon delving.

Personally, I am a fan of shorter, more focused dungeons. Overland exploration is my favourite part of the game, so I make dungeons short, deadly and sweet.


I run big dungeons, often generating them on the fly. Good rooms often have barriers (“liminal spaces”) instead of just doors. Once you figure out how to get through, around, or over the barrier, you can add this room to their “controlled space”. It’s assumed that adventurers move at a snails pace through rooms they haven’t explored yet, and very quickly through ones that they control.

You mention that the interesting parts of D&D are “rather improbable” in a dungeon. Consider: fungal biomes, dungeon merchants, subterranean wars and ruins, and the absolute freakishness of just… walking into a completely ordinary tavern, however many fathoms below ground.

Treating dungeons as dry and dead caves is “realistic” sure, but kind of boring. Treating it as a mythic underworld (page 22) is more interesting. It doesn’t have to make sense in the that physics demands, but it can be self-consistent in a new way. And of course, there’s Veins of the Earth and the Veinscrawl if you want to go hard on the faction side of things.


I’ll echo Spwack’s point above. Also, I’m currently seven sessions deep (last session last night) into running Castle Xyntillan, which has almost 300 keyed rooms, a few empty ones I fill up, and a nearby town. At this point the party has explored maybe 35-40 rooms, and haven’t come close to uncovering the town’s secrets. They’ve spent a good part of their time in town, maybe a two sessions totaled up, but the vast majority of play has been in the dungeon. My sessions also usually go between 3-3.5 hours, with 10-20 minute max combat less than once a session(they either win fast or run away). They spend much more time than that interacting with NPCs back in town.

It’s worth noting that CX, like the best of the large dungeons from previous years, is very colorful and varied. There’s no point where you move from room to room and can predict what you’ll find behind the next door, which keeps it all feeling fresh. There’s also a ton of NPCs, and with reaction rolls most of their interactions with the party end up being non-violent.

The pressing reason to keep delving at first is always loot, and the fact that, at least in my experience, players need little excuse to explore and throw their characters into danger. Over time, the campaign has developed some ‘quests’ mostly by taking the elements that the party was intrigued by and building on them as I go. They’re currently getting ready for a wedding between an NPC adventurer and an undead princess, while in a battle of wits with a vampiress, who is the maid of honor (I never said my campaign was serious).

That kind of player-centric emergent play is what megadungeons excel at creating if you set them up right, as opposed to small dungeons which provide a more contained and curated experience. Both may be dungeons, but the difference in scope plus the inclusion of factions and interactions over the course of multiple expeditions create a totally different experience to the small dungeon.

TL;DR the best big/megadungeons include those other pillars of play within themselves and maintain a great deal of internal variety to ward off monotony.

(Off topic, but I’ve never run interactions with more than 1 or 2 NPCs at a single time before. It just hasn’t come up in my short GMing career.)


I have run several large dungeons (the largest of them being Rappan Athuk, Barrowmaze, and Dwimmermount), and like things in general, it’s not for everyone. One of my players loves mapping and is incredible at it, while another can hardly tolerate the slow pace that sometimes is quite necessary for dungeoneering.

Specifically to your two points:

  1. There always needs to be a reason going forward. This might be an external goal (like a quest/geas), but in my experience the best reason for pressing on is genuine curiosity. When I’m a player, I want to know what’s behind that corridor or what’s the deal with this thing I heard rumours about. Of course, when I’m actually there, I might chicken out, because survival is the most important thing, but at least I had to venture deeper to discern whether I actually want to engage with the specific thing or not. Bottom line: if there’s no internal or external motivation for exploration, there won’t be any.

  2. There certainly are a lot of fun things to do in a dungeon, like playing factions off each other, interrogating denizens (either captured alive, charmed, or via speak with dead/plants/animals - but some of them are friendly anyways), figuring out puzzles, avoiding or dying to traps, etc. Heck, a literal underground tavern/inn, undead theatre, or orcish gambling den are all things that could be found down there. Of course, these things are going to be the spice not the meat; if you don’t enjoy dungeoneering that much or you want every session to be more varied (a goal I respect but find not feasible or even desirable at my table), then of course smaller dungeons and lairs are just better for your purposes.

That such a small number of rooms and a single combat might take hours, as you mentioned in the intro paragraph, I find surprising. In my experience, players can easily map 10-15 rooms in a few hours with a handful of large fights, puzzle solving, and the like sprinkled in. Granted, I only describe things briefly, especially in combat - but when having a party of 4 PCs + 2 classed retainers + 6 hirelings fighting 2 ogres and 12 orcs is expected to happen, quick decisions and speed trump almost everything.

(For that end, it’s good practice to allow marching three abreast - I believe in by-the-book B/X that is only feasible for characters armed with spears -, so that even in choke points, such as doorways and corridors, 6 melee blows may be exchanged each round. At least that’s how we’ve been doing it in Melan’s current campaign, and it works really well.)


I often have trouble running combats quickly, even using lightweight systems like Knave (maybe this is a result of coming from a video game background that led me to D&D 3E and 4E). How do I fit a handful of large fights among all the other dungeon crawling stuff in only a few hours?


Ran into this a handful of times. What really help me was Index Card RPG’s approach to room difficulty ratings. Skip to around 9:30 of this Dungeon Craft video to see an example:

This helped me streamline entering combat, handling combat stats during the encounter, and adjusting difficulties (I try to avoid this however) on the spot. If you’ve identified that a room might take longer than maybe you’d like due to the difficulty score, just adjust it on the spot to save time.


Assuming side-based initiative, the way we play is this:

  • Both sides roll 1d6. Higher goes first.
  • Characters act in marching order. So, the first line goes first, then the second, then the third, etc.
  • If it’s a planned encounter (in the sense that the monsters were keyed to the location), then I have pre-rolled hit points for each of them. If it’s a random encounter, I roll their hit points upon receiving damage.
  • When there’s a question regarding the environment, I answer it honestly but also as simply as possible. Brevity is key. Same goes for mechanical resolutions in case of tactics not covered by the rules. It helps if the rules are easy to extrapolate (in Dungeons & Companies or Helvéczia, for instance, most unusual manoeuvres are opposed attack rolls that succeed on beating the opponent by 5 or more, or maybe 10 or more for really difficult stunts - that’s it).
  • As mentioned previously, the more members on both side get to engage, the faster combat goes (so if even in a doorway or on a corridor it’s 3v3 in melee, that’s going to accelerate combat).
  • As the GM, you have to make quick decisions for your monsters all the time. However, you ought to urge your players to do the same.
  • Check morale on first casualty, then again after losing half of forces. Also check it if something terrifying happens (like a particularly bloody dismembering of an enemy, or when one side is seemingly impervious to damage, etc.).

Hm, I’m very strongly against changing difficulty on the spot for any reason.
At any rate, this seems like a sensible way to simplify things if there’s a need to do that.

1 Like

Would agree in general - I’d encourage anyone to avoid doing this but I have definitely done it when it’s needed. Usually due to time constraints of my players. One thing I should clarify though is I would almost never do this with the mindset of “this is too hard for the players” - I only adjust difficulties when I can easily identify that the encounter, as whole, really wouldn’t be a large threat in the first place, so adjusting the difficulty for time’s sake is a bit of an easier sell to me at that point.


I’d recommend Reading maze of the blue Medusa.


I have, but it’s not really my jam aesthetically.


There are some impressive precreated dungeons full of differences from area to are but I also think that this problem is solved for me by making twenty or so rooms before session one and then only making the new rooms of a large dungeon after the first session and between all the others. That allows the dungeon master to make rooms that follow the mood that is created in the first session or extend discoveries that become good in play


It’s not your job to adjust the difficulty of a fight. The players can do that, either by being better at the game, or running away.

Everybody’s table is unique, I don’t quite think this steadfastly when it comes to my “job” at my table.

This is valid. My comments were through the lens of saving IRL time based on time constraints however. Me changing the room difficulty to save session time before the party is even aware of the room’s contents doesn’t disparage the player’s IRL skill level with the game or ability to create alternate “out” solutions in character such as running away, in my opinion. I rarely do this, and again would discourage it generally, but it is something I’ve done to speed up larger dungeons while working with my player’s time constraints.

1 Like

Oh that makes way more sense. Yes, there is absolutely no resource more scarce than IRL time, and anything that can be done to save time that way is worth doing! Especially if you can do it before the players actually “arrive”, which is a skill I have yet to master

1 Like

I am happy that this thread yielded so many responses and points of view! :slight_smile:

I have to say, the more I play, the more I feel almost anything at all the makes the sessions more fun is worth trying at least. And people often forget (not you Spwack!), it has to be fun for the GM as well, and if that means changing things, even if it is not considered “the right way” by some, is fair game.

I have come to the point that, if I really don’t gel with a player (often times because they are mildly dickish and/or have a very different expectation or play style),… I kick them out. In as nice a way as possible of course! But I don’t want to end up in RPGhorrorstories about how a player is ruining my game or my group is going to hell. I am prepping, hosting and running the game, if it simply doesn’t fit well together, best to nip it in the bud. If I am not having fun, odds are all the players will notice and have less fun, this can easily become a cycle.

Saving time is just below “make sure everyone (me included) has the most fun possible” in importance. It is too easy to not fave fun, or get GM burnout, or even start dreading DnD sessions a little if time becomes ever more scarce, incrementally.

I noticed that is the one thing that would kill my motivation or fun, if prep or dealing with a certain player started to cost more and more worries or time, or the game started bogging down and slowing down more and more (rules etc) that was a sure fire recipe for me enjoying it less and less. Which more or less brings it back to my original question. Less = more for me. But it is different for every GM and the many approaches in this thread and links did give me good ideas to expand what actually works for me! :slight_smile:


As long as the dungeon actually makes sense, treat it like a village or town, with locations that are high traffic and locations that are less busy. Commit to memory the basics of the description of the place as a whole and what sense information the player character will have - hot or cold, stinks or not much smell, and so on.

Then read the dungeon as written to see if it makes sense. Understand the layout and why it is laid out that way- if there is any real reason for the layout. Surprisingly often dungeon maps are complete nonsense. Think about line of sight and hearing things.

If it’s a railroad type dungeon that is, once you’re in it’s hard to get out (why?) then the layout is crucial. If it’s more like a real life dungeon tunnel complex, it is about going in and retreating, going in further, and so on.

There is a tiresome quality of artificiality to many dungeons that strips them of any real interest, that is the biggest problem to overcome.

Operation Unfathomable is the best Underworld dungeons I’ve seen. It has roads between zones. You’re not just seeing wandering monsters, this is a biome for weird animals. And rival cults and merchants have parties and processions which travel around often.

The Anomalous Subsurface Environment took some cues from Thunder and has some pretty surprising villains. It has factions and zones.
It has dungeon elementals who reset traps which are a sort of boogyman. It could be interesting, first to see one in the corner of your eye, then see one full on before it vanishes into a wall.
And then it has rooms which point to a mysterious organization which conducted experiments here, inspired by the series Lost.