Traps Are Not a Tax on Movement
The worst trap in D&D is one that just does damage. You step on a pressure plate, you take 2d10 piercing, you move on. That’s not a trap, that’s a toll booth. The version worth building is something that makes the party stop, look at each other, and figure out what to do next.
Think About What the Trap Is Actually Doing
Traps operate on four components: severity, trigger, effect, and countermeasure. The severity gets calibrated against character level using save DCs and attack bonuses, but that’s the least interesting part of the design. The question worth asking first is what role this trap plays in the dungeon’s ecosystem.
A simple alarm trap that does no damage can be more punishing than a blade pendulum that deals 4d10. If the alarm wakes every cultist on the floor, the trap hasn’t hurt the party directly. It has just made the next three encounters significantly worse. That’s a good trap.
The trigger is where most DMs underthink things. It should reflect who built this place and why. A magical fortress staffed by wizards might use an alarm spell keyed to certain verbal triggers, with a password clue buried somewhere earlier in the dungeon. That’s fair. That’s even fun. A pit trap under a random hallway in a king’s banquet hall is neither of those things.
The Disarm Is a Puzzle, Not a Roll
Mechanical traps have physical components. If you build one with an actual working logic in your head, like a peg holding a counterweight that holds a portcullis, then players who reason through it correctly should just solve it. No Dexterity check required. Cut the rope, the stone drops harmlessly because there’s nothing for it to pull anymore. Done.
This requires more prep, but the payoff is enormous. Players who feel like the world makes physical sense trust it more. They engage more creatively. Draw a simple diagram if things get spatial. It doesn’t have to be good. It just has to get everyone on the same page about where the lever is relative to the door.
Passive Wisdom (Perception) is genuinely what it’s for here. If a character’s passive score beats the trap’s DC, give them a hint in your description. A glint. An odd shadow. The faint smell of something off. Let them discover it themselves rather than announcing “you detect a trap.”
Complex Traps and Initiative Order
Complex traps run on initiative and evolve over multiple turns. The Sphere of Crushing Doom from Xanathar’s Guide is the clearest example: a giant steel sphere rolling down a 150-foot sloped hallway, looping through portals at each end, gaining speed and damage every round. It starts at initiative 10. Once the damage climbs to 10d10, it moves at both 20 and 10. The party has to dispel the portals, stop the sphere with something like wall of force, or find whatever object from earlier in the dungeon the DM seeded for exactly this moment.
The design principle is that the longer it runs, the worse it gets. That’s the whole structure. And it works because every turn is a decision, not just a saving throw.
One final thing: don’t overuse traps. A dungeon where every door and floor tile might be trapped doesn’t feel dangerous, it feels annoying. Players will check every surface, every step, and the game grinds to a stop. Traps work because they’re unexpected. Use that sparingly.
For reference, simple and complex trap design appears in the Dungeon Master’s Guide and Xanathar’s Guide to Everything.