log in or register to remove this ad


5E What Unearthed Arcana Subclasses Would You Allow in a Campaign?


I'm planning to run a new campaign eventually and thought it might be fun to let people use certain Unearthed Arcana options (this could also hypothetically include other sources, like the Domains from Plane Shift: Amonkhet).

Which ones are generally considered the best and most balanced? I've personally only ever played a Way of Tranquility Monk.

log in or register to remove this ad


Goblin Queen
I allow any options in the player’s handbook by default, and will specifically mention any exceptions or additions to that for a specific campaign. If a player wants something outside of that, I ask them to talk to me about it specifically. I’m usually fine with allowing UA options, but not before discussing it and how it will fit into the campaign.


Current in-testing Unearthed Arcana options are allowed by my books.

I also tentatively allow certain options from DMs Guild, En5ider, and D&D Beyond, but those are really on a review basis. Even beyond balance concerns, I don't like the idea of overlapping subclass themes within the same class. If it's an archetype of the class, it should be able to stand out from the crowd and be its own thing. Death & Grave (and also Twilight & Trickery) Domains are already too close for comfort before allowing things like Entropy Domain from Ruins of Mezro & Xanathar's Lost Notes to Everything Else, especially now that we also have UA options for swapping out Divine Strike for Potent Cantrips in the Variant Class Features playtest.

Some classes I'm more courteous with because they just don't have enough options in RAW+UA - say Artificer or Ranger (but again, a lot of those options published are working off of bad versions of those base classes, like the PHB Ranger or the 2018 Artificer). I'll allow PHB Ranger, mind you, but I strongly urge folks to use the Variant Class Features Ranger instead.

Sir Brennen

Pretty open to allowing them, though I am kind of regretting letting a known min-maxer make a Mystic.
Last edited:


This. They need to be playtested somehow.
If you don't playtest them, how are you supposed to vote on them in the survey?
Granted, the surveys are usually not around long enough to do proper playtesting.
Usually not enough time to playtest effectively before the survey turns around (unless you play weekly, maybe), so I gather most of us give feedback based on theorycrafting. Early on in 5e, they specifically requested that we only give feedback on classes we've sat down with at a table and tried in practice. They don't make that request anymore, because they know the turn-around time is too quick, it's too hard to control (can't check if someone is telling the truth or not), and they get so much feedback, positive and negative regardless, that the public UA playtests (outside of the long DMsGuild periods for Artificer and Mystic) are really more for the purposes of seeing what sticks or not with a plurality of players. Would we play as this if it was in a book?

That's more important that the exact lessons we learn from playing with it right now. If we have the time to do so and then give feedback, that's even better. But they get a lot of feedback, and we can learn a lot from thinking about these rules and discussing them with each other here on the forums whenever a new one is posted (we, of course, are also a minority of feedback-issuers, but there are a lot of people theorycrafting elsewhere, too).

Probably not the Lore Wizard. Possibly not the Mystic without a good run-down of the character's capabilities. I would probably at least insist on reading the UA before I made a decision on an option I was unfamiliar with.

Also, if there is a published version of an option that appeared in a UA earlier, I would probably say stick to the most recent or official version unless there were significant differences between them.

Li Shenron

One of our PCs is a Swashbuckler. I should eventually buy XGtE but in the meantime we've been using the UA version.

Another PC is a Centaur. Originally it just used the statistics for a human character. We ret-conned it later against the UA version but not wanting to really recalculate anything or give the PC a net advance, we only added the Hooves weapon (ribbon) and the Equine Build property (encumbrance bonus, climb penalty).

I am not generally looking at UA as stable/usable material, but if something comes up that the players wants to play that is not explicitly available in the PHB, and if I recall some UA article covered it, then I can bring it up. My players usually aren't aware of the existence of UA, so they don't come up with "can I use <this> from UA?" requests. The Swashbuckler was my idea when the Rogue reached 3rd level because the player was a bit skeptic of criminally-sounding archetypes (Thief and Assassin) and had no interest in magical abilities (Arcane Trickster). Centaur was a player's request for purely narrative reasons.
Last edited:


I'd allow pretty much anything, using later versions if there has been a revision. Still have that psychic artificer subclass in a game, thinking of altering it so that it works similar to current artificer subclasses because the invulnerable psychic brain thing is really annoying and I feel it should be able to take damage.


Mystic is really the only one that's flat-out no go, which is sad because I really want a psionics class. I'm pretty easy going in regards to homebrew, with the caveat that anything you ask for I will nerf and/or rewrite to fit the balance.

Traditionally I haven't allowed them at all. However, likely for my next campaign I'll at least consider ones that are thematically acceptable to the campaign, and that aren't mechanically problematic (not happy with ANY of the psionic choices, for example).


I tell my players they can ask to use Unearthed Arcana material. I generally allow it, but evaluate it before allowing it. The same is true of 3rd party materials.

COMING SOON: 5 Plug-In Settlements for your 5E Game