What skills/stats are used for Lockpicking, Finding Traps, and Disarming Traps?  I know you use you Thief's Tools for lockpicking, but is that a Dexterity roll, +proficiency bonus if you're proficient with the tools?  Or an Intelligence roll?

What about finding traps? Wisdom (Perception)? Intelligence (Investigation)?

Does Disarming Traps use Dexterity (Sleight of Hand)?  Does that require Thief's Tools, and if so, does it use the tools proficiency rather than Sleight of Hand?

Btw, I'm talking about standard locks and traps... I'm sure the DM can put locks and traps that use non-standard rolls (say, a lock with fifty different buttons, which requires the picker to notice which one is ever-so-slightly different (Perception) to press it in order to unlock it).

1st. I recommend Thieves' Tool checks to unlock a standard lock should be Thieves Tools (Dexterity).  My brother is a bonded locksmith, he said it's all about dexterity.

If you're trying to do something more interesting then change it up.

As for finding traps.  I recommend as a GM not saying you "have to use a particular method".
Most of the time use Perception (Wis) to detect most traps, because it will be about spotting something and it's what most players expect.  On the otherhand is a player says "I want to use Investigation to look for a trap" that should be encouraged.  Don't make them guess what is in your head as a GM.  An investigation check would just be "if I was going to make a trap..."

Major thing.  Do NOT make the trap disarmer put lots of skills into doing this.

I know a lot of people now use Perception (Wisdom) to detect a trap, Investigation (Int) to decide how the trap works, and Thieves' Tools (Dex) to disable it.
This is terrible.
1) It requires the trap disarmer to take proficiency in 2 skills AND a tool set and have good ability mods in 3 different score.  This is MAD as hell, and awful.  Adding Sleight on Hand (Dex) onto that just adds another layer.
2) The more rolls you make a player make the more you guarantee failure of any particular roll.
3) Nothing else works like this.

You have to ask yourself.  How does the challenge of a lock or trap improve play?
If a trap is merely wandering damage, the players are going to get annoyed by them and become paranoid.  They will open every chest from 30' with Mage Hand or Thaumaturgy.  Declare they are standing *next* to the door not in front as they open it, etc... etc....

And remember as the GM you describe the situation, if you describe a featureless hallway do you want them moving 5' and making a perception check every turn?  That doesn't help anyone.  If you want them to be on the look out and have the trap be an encounter.  You have to describe the situation as such to give them hints.

FullMetalBunny: Well said. 

Thank you for the detailed analysis/suggestions!

I wasn't thinking of using multiple rolls in order to open a locked, trapped chest, like you describe.  I was mostly wondering whether there was a standard method for these.

That is, say the party encounters a locked door, and they want to pick the lock.  I'm quite sure that uses Thieves' Tools (and hence the roll gets proficiency bonus if the picker is proficient with them), but does it use a stat bonus, also (perhaps Dexterity)?  Or say the party finds a chest, and they want to make sure there are no traps.  Is that an Investigation check? Perception? Either? Does it depend on the trap?  And let's say the chest is trapped, and they find the trap... what do they roll to disarm it?

Also, just to be clear, I'm asking this in this forum (Rules & Game Mechanics), and not in the DM's forum, because I'm curious as to whether there's a standard/official answer (which, I admit, were I not too lazy, I could look up myself... but being at work, checking and posting on a forum is less conspicuous than poring through the rules manuals :D).  If/when the situation comes up while I'm DM'ing, if I don't have a standard/official solution, I wouldn't have a problem coming up with a satisfactory solution.

Which is not to say I don't greatly appreciate your input (I do!), tho!

• The trap rules suggest: "A character actively looking for a trap can attempt a Wisdom (Perception) check against the trap’s DC" and "You might call for an Intelligence (Investigation) check for a character to deduce what needs to be done, followed by a Dexterity check using thieves’ tools to perform the necessary sabotage".
• I instead use: Perception or Investigation (player's choice) to notice the trap, followed by a thieves tool check to disarm.
• However, for magical traps I use the suggested: "Any character can attempt an Intelligence (Arcana) check to detect or disarm a magic trap"

I believe Thieves Tools were designed to be used to disable traps.  "Proficiency with these tools lets you add your proficiency bonus to any ability checks you make to disarm traps or open locks" https://www.dndbeyond.com/equipment/thieves-tools
As for what Attribute to use with Thieves Tools.  There is no "official answer", unlike Stealth, Perception, Athletics, etc...

I do know Dex is the default one people use, just as Cha is usually used for playing an instrument.

As for do you use Investigation or Perception to determine if the chest is trapped.  There isn't an "official" answer, it's your call as the GM.
There is a passionate discussion here: https://www.dndbeyond.com/forums/dungeons-dragons-discussion/rules-game-mechanics/2117-perception-vs-inspection-check

The RAW rules literally say both skills can be used to detect hidden things.
https://www.dndbeyond.com/compendium/rules/basic-rules/using-ability-scores#Investigation  "You might deduce the location of a hidden object"
https://www.dndbeyond.com/compendium/rules/basic-rules/using-ability-scores#Perception "When your character searches for a hidden object such as a secret door or a trap, the DM typically asks you to make a Wisdom (Perception) check."

My GM hat is let the player who the skill they are best at.

Perfect, thank you!

 fwiw: you probably don't even need to have them ever roll Perception. Passive checks "represent the average result for a task done repeatedly, such as searching for secret doors over and over again, or can be used when the GM wants to secretly determine whether the characters succeed". If your group's standard procedure is to constantly search everything, and they always assist each other, just use the highest passive perception (+5, to represent the assist).

As MattV said, consider using the passive skills of the character first and only if the skill can't match the DC then ask for a roll.

For example, if there is a trap that requires a DC15 perception check to notice, a DC12 investigation check to figure out how to disarm it and a DC14 check to disarm it ... then consider using the passive perception check to find the trap, the passive investigation check to figure out what is required (assuming the character is actively looking for traps, there is no special time pressure and no consequences for failure the character SHOULD be able to use the average result represented by the passive check to resolve noticing the trap and figuring out how to disarm it). If disarming the trap has no consequences for the characters (e.g. a pit trap where once you are aware of it you won't fall in it) then maybe even use the passive thieves tools check for disarming/disabling the trap. If the passive skills aren't up to the task then allow the character to roll to see if they can do better than the passive result.

Basically, in my opinion, it is worthwhile using the passive skills to prevent the dice rolls from getting in the way of the narration of what is happening in the game. There are a lot of checks that can be resolved by passive applications of skills and this in turn lets the character who has invested in those skills succeed ... rather than having the 8 int barbarian who rolls a 20 on an arcana check succeed on a DC12 task while a 18 int wizard who rolls badly fails on a check where they are proficient.

Certainly, depending on the circumstances, passive checks can suffice.  Like I said, I wasn't really asking for DM advice, but rather an rules question.

On the other hand, I disagree with the "character who invests in a skill should succeed while character who is bad at it shouldn't" concept.  Or rather, I disagree with it as a general rule.  There's a reason we roll dice to determine outcomes while playing D&D.  And often, it makes for a fun, maybe even more fun, story.  As an example, recently, there was a bit of religious lore the characters in the game I'm DM'ing could recall which would help them progress.  The druid rolled low, the wizard (not particularly religious, but with a high Intelligence, so a decent Religion bonus) also rolled low, but lo and behold, the Half Orc Barbarian, not proficient in Religion, rolled very high.  So it turns out he remembered a bedtime story his human mother used to tell him, about how Torm would always watch over him...

At lower levels, when the relevant bonuses are smallish, the difference between the skilled and the unskilled isn't very noticeable, so it might seem unfair that the high-Charisma Paladin isn't much better at Persuasion than the gruff, loner Ranger, but eventually the difference will be more marked (if the Paladin, in fact, is proficient at Persuasion, and the Ranger is not), and you have to consider that at low levels, the characters are still not masters of their craft.  A novice Rogue can, and will, fail to pick locks fairly often, while a master Rogue can make it so they'll never (literally) fail to pick most locks.

How do you guys handle failed lock picking attempts?  Do you allow infinite retries (which essentially makes the DC of the lock irrelevant unless you're doing it in combat)?  Just allow one attempt (which seems too punitive)?

I've been allowing 3 attempts, with the DC ticking up by 2 each time.  By the third failure, I figure the lock is too dorfed up by the character's fumbling to be openable (other than by force).

Here is my thing about "not allowing reattempts"

It's a lock.  It's a machine, designed to be resistant.  Anyone who has actually done lock picking knows it can take a while.  A check is 6 seconds of work.  It's why I miss the "Take 10 and Take 20" Rules from 3.5.  My default back then was to roll a single check, then declare Take a 10.

If a lock can be messed up by a pick then a key is going to jack it up hard!

And yes, a lot of the time you can stick in a bump key, hit it with a hammer and be in... but that's loud and only works on Home Depot locks.

So, you allow infinite reattempts?  While that may or may not be realistic, from a gameplay perspective, it would seem to make locks kind of meaningless unless there's time pressure of some kind.

I use Perception to notice it (usually passive), and a passive investigation to see if they instinctively deduce how it works and how to disarm it. If they do, the trap's disarm DC is lowered. If they don't, they can either figure it out by the description I give them (I don't just say 'you find a trap' I describe what's there, they can disarm it or avoid it at varying DCs) and decide HOW they're doing it, or they can just make a thieves' tools (Dexterity based) check against the standard DC. Most Rogues will be fine with the latter method, but against particularly ruthless traps there's a little more caution used. 

As an addendum on the topic repeated/retried checks. 

If there is NO penalty to the roll, no time crunch, no negative outcome of a failure -- there is no tension in the roll, and there is no point to the roll in the context of making a fun game or story. It is momentum death to put a locked door in an empty room and ask the Rogue to make checks again and again until they hit the magical DC you've set. Ugh. Any Rogue worth their salt will eventually pick that lock given enough time. So if you're going to have them roll, just do it to determine how long it takes them. There's no 'failure' here, it's just 'this took you a few seconds, this took you a few minutes, your party was groaning as you worked at the lock for a good fifteen minutes'. 

There IS penalty to disarming a trap and failing. Why would the trap not go off? That's the penalty for failure. The trap goes off, damage is taken, or negative effects occur. If the trap doesn't reset itself, well now it's gone off - what happens? Either there's a new problem (fire blocking the path! water flooding the cavern! an alarm went off to notify the enemy!), or the party is worse for wear, or the trap resets... And now a reroll/retry makes sense. You're now adding a penalty to failure.

This failure naturally exists in some rolls, which should also not be repeated ad nauseum. A character who fails on a History check to recall the lore on this gem-encrusted fork doesn't know. It doesn't matter if they roll again, they don't know -- they aren't magically going to remember now. A character who fails a Deception check to trick the noble into believing she's rich has failed, if you roll again your lie isn't suddenly believed. 

If you're going to ask for a roll there should be a clear benefit, and a clear detriment. What is gained through this? What is lost if you fail? How does the story's trajectory change when the Rogue fails to pick this lock? Does the party become injured, do they break their tools and can't use them for the rest of the dungeon, does the enemy know their location, do they lose time in their escape? If you're ruling it that they simply can't get through the door now -- have you written yourself into a corner? Is there somewhere else they can go? 

Personally as a DM I never ask for a check more than once. You either succeed, or you fail, and the story is altered and moves on based on that. If there's no way for the story to shift on a failure, there's no point for a roll as long as someone is proficient in the skill. 

Quote from Generic_Poster >>

So, you allow infinite reattempts?  While that may or may not be realistic, from a gameplay perspective, it would seem to make locks kind of meaningless unless there's time pressure of some kind.

 Yes. Exactly, that's what locks do.  They are a barrier based on time and/or noise.

If you haven't created a scenario where time and/or noise are important issues, then that's a problem you as the GM should create.  Don't griefer your players with locks. 

Take the example of the PCs want to rob a store, because it's run by the Dragon Cult and there they know there are documents they want to steal.
The PCs have a few options.

1) They can try to steal the key off of the shopkeeper.
2)The Fighter takes a sledgehammer and makes an Athletics check to break the door down, and the party races against time to find the documents.
3) The Rogue can try to sneak up to the door in the dark and pick the lock (lockpicking).  Now here time is going to matter, because someone will walk by and there is no place to "hide" on a city street.  They are designed for that purpose.  The Rogue can't take Passive Lockpicking because a guard or citizen will walk by and either stop the Rogue or call the Guards.  You as the GM can decide how many chances the Rogue has to pick the lock before she needs to try something else.
3a) IF the Rogue can't get the lock, but the guards were not alerted.  Maybe the Party blows a 2nd level spell on Invisibility.  The lock is now bypassed with time at the cost of a spell slot.
3b) IF the Rogue gets the lock and the Rogue/Party have Darkvision then any further locks inside will be pointless.  You might as well just describe them because there is no time pressure and the rogue will bypass them.   Ask for a check to describe if it's easy or takes a long time to get past future locks.  This assumes they can find anything because with Darkvision all Perception is at DisAdvantage.  Maybe they can't find what they are looking for without light.  They need to light a lamp then it's possible someone on the street's neighborhood watch will eventually notice.  Suddenly time is important again, it's a race against the watch investigating.
3c)  IF the Rogue gets the lock and the Rogue/Party have Darkvision then any further locks inside will be pointless.  In your mind the party is taking too long.  Have an errand boy come with a late night delivery or the shop keeper's daughter (she's probably an innocent, so they are unlikely to kill her) is sent late at night to get something that was forgotten.

MellieDM, I disagree with you about the History Check.
In the moment the character is holding the gem encrusted fork, they flub the test and don't know anything.  Just failing the check creates no tension.  Wait until the fork is not around anymore and use l'esprit d'escalier.  Have the character with a high History Score, who flubbed their roll remember something about the fork now that's it's too late.  That's how you make Passive Checks interesting, the tid-tip gets dropped but it's probably to late to do anything now, but it will be known for the future.

I do the same with Nature/Arcana/History check on monsters.  When the monster is barrelling down on you, you get 1 chance to remember something important about it.  If the party knows they are dealing with a Rakshasa and planning ahead.  Time isn't an issue.  Give them information based on their best Passive Check.  If two people are proficient give it Advantage because they are obviously working together, and if that check is high enough there is little reason they wouldn't know Piercing magical weapons wielded by Good Creatures is how you defeat them.

This is the best answer ever.  LOVE IT!!