How do I run secret doors in D&D without endless Perception checks?

One door, three different decisions

Keep detection and opening separate. A successful passive or active detection locates the door; the clue can leave its catch unexplained. Opening is a further problem only when the mechanism earns that status, and a missed optional door can stay hidden.

Noticing: a passive success locates the door

In the 2014 fifth edition rules, the Dungeon Master's Guide uses passive Wisdom (Perception) to decide whether a character notices a secret door without actively searching, and an active check for a deliberate search. A passive score is 10 plus the character's Perception bonus, adjusted for advantage or disadvantage where it applies. When the score meets the door's detection DC, locate it for the players with a concrete sensory description. The catch can remain unknown:

• A draught where there should be none Cold air moves across ankles or bends a candle flame at the carved panel. The players know which part of the wall hides the door; the draught does not explain its catch.
• Wear that does not match the room Scrape marks arc around one panel, its mortar sits slightly proud or recessed, or dust is thin where hands keep touching stone. The wear locates the door without giving away how it opens.
• Sound and measure A panel sounds hollow when tapped, or measuring shows the room is shorter inside than the corridor outside suggests. These clues locate the door; tapping and measuring can also reveal it outright when the action settles the question.
• Check the senses before you award the clue Ask whether this character could plausibly notice it from here: light or darkvision, distance from the wall, noise and distraction, and whether they are hurrying or keeping watch for threats. A high score held by a character at the back of the party in pitch darkness has not automatically perceived the hidden door.
• A missed clue is not a locked scene If no passive score meets the DC and nobody searches, say nothing specific and move on. An optional door can simply stay hidden. A door the adventure needs is a design problem, covered below, not a reason to prompt roll after roll until someone succeeds.

Searching: roll only if the door remains undetected

Do not ask for another detection check after a passive success has located the door. If passive detection failed, or the players have a reason of their own to suspect a wall, hear their approach first. Call for an active check only when it could plausibly fail, the result is uncertain, and failure would cost something.

• No roll when the action settles it If a player holds a flame along the seam, puffs flour or chalk dust at the crack, pours water and watches where it runs, or taps stone by stone and listens, and that action would reveal the door, it reveals the door. Dice resolve uncertainty; they do not replace a decision the players have already made well.
• One lead search, with real help Let the best-placed character lead. In 2014, another character can help when they could attempt the task alone and their cooperation is productive, such as holding light steady or listening at the other end. In 2024, the helper must choose one of their own skill or tool proficiencies, and the check uses that proficiency; practical assistance alone does not qualify.
• Distinct jobs are not a dogpile There is a difference between four characters rolling Perception at the same wall and a search with divided labour: one listens, one measures the room, one checks the floor for wear, one watches the corridor. Give each distinct job its own consequence if it has one, or fold the help into the lead check. Identical rerolls after a failure add dice without adding play.
• Identical retries need a change If nothing has changed, the same search gets the same answer, so do not ask for the roll again. A new roll becomes fair when the approach changes, better tools or light arrive, new information turns up, or the party accepts a cost such as more time, more noise or a resource spent.
• Failure moves the scene A thorough search takes time whether it succeeds or fails. Let that time advance circumstances already in motion: the patrol on its regular round moves closer, for example. If time is unlimited, nobody is watching, and nothing changes during a slow search, consider letting a thorough search succeed without a roll; a check with no cost for failure only decides how long success takes.
• The ten-minute search is a house rule Some tables use a house rule that carefully searching one 10-foot section of wall takes ten minutes and finds a door there if one exists. That is a workable house rule, not a universal fifth edition rule. If you adopt it, say so plainly and let the cost be the time itself and whatever that time brings with it.

Secret door decisions at a glance

Opening: a second problem only when it earns its place

Discovery, understanding and opening can be three different tasks, but they do not have to be. The 2014 Dungeon Master's Guide makes the same distinction: once a secret door is detected, an Intelligence (Investigation) check might be required to work out how to open it if the mechanism is not obvious. The word to hold on to is might. A sliding panel with a worn thumb catch opens. A door held by a counterweight, a sequence of studs, a key kept elsewhere, a magical seal or a trap is a separate challenge because the fiction makes it one.

The 2024 revision keeps passive Perception and gives deliberate searching clearer action labels: Search for discerning something that is not obvious, commonly with Perception, and Study for deducing how something works, commonly with Investigation. The labels change; the habit does not. Do not add a redundant Investigation roll as a default tax on a door the party has already found fairly, and if opening does need a roll, decide in advance what failure does: jam the mechanism, make noise, take time, or trigger the seal, rather than leaving the party staring at a found door they cannot interact with.

Reward tools and preparation here as much as at the search stage. A crowbar, chalk, oil, a ten-foot pole, a familiar sent through the gap, or a rubbing of the stud pattern taken away to study can each change an uncertain opening into a settled one, or turn a failed attempt into information about how the mechanism resists.

Essential doors and optional doors are different designs

An optional secret door can be missed. A vault of extra treasure, a shortcut that skips a guarded hall, or a shrine with lore rewards the observant and stays shut for everyone else; that is the reward working as intended, and the missed door costs the party something they never knew they had, which is a fair price for not searching.

Never gate essential progress behind one easily failed check. If the only route to the next part of the adventure sits behind a DC 15 search, a bad roll has not created tension; it has ended the adventure. Give an essential door at least one backstop you prepare in advance: a second clue in another room, an occupant who uses the door where the party can witness it, a longer and more dangerous route around, a way to force it at a noisy cost, or information elsewhere that sends the party back to that wall with a new method. The backstop can be slower, louder or more expensive. It only has to exist.

Before you hide a secret door

• Write down the clue in concrete terms: what a character can see, feel, hear or measure, and where it can be noticed from.
• Note the passive detection DC: a successful score locates the door; the catch can remain unknown.
• If passive detection fails, decide what might give the party an independent reason to search, and which described actions simply work without a roll.
• Set an active check DC only if that search could still fail, and note how its time advances existing circumstances or creates a concrete cost.
• Rule on help before the roll: who leads, whether the helper could attempt the task alone (2014), and which skill or tool proficiency they choose (2024).
• Decide whether opening is automatic once found or a distinct problem, and only choose the second if the mechanism, key, seal or trap makes it interesting.
• Classify the door as optional or essential. If it is essential, prepare an alternate clue, route or costly recovery before the session, not during the stall.
• If you use a timed thorough search that always succeeds, label it as a house rule at your table and track what the spent time changes.