How to Split Loot in D&D: 6 Methods (and When to Use Each)
Six fair ways to divide D&D treasure, from need before greed to a self-balancing loot auction, each tested on the same hoard so you can pick one tonight.
Your party finally cracks the vault, and on the table sits the kind of haul that should feel like a reward: a +1 longsword, a Cloak of Elvenkind, three potions of healing, a wand of magic missiles, and 1,800 gp. Four of you, a fighter, a rogue, a cleric, and a wizard, lean in. Within about thirty seconds the longsword has stalled the whole game. Everybody wants it, nobody agrees who should have it, and the session clock keeps ticking while you talk in circles about damage dice and who carried the fight in the last room. The hoard is great. That isn’t the problem. The problem is that the moment it landed, you discovered your table had never actually decided how it splits things, and a four-way standoff over one magic sword is a miserable way to find that out.
I’ve watched this happen at more tables than I can count, and the loot is almost never the reason it goes sideways. The contested drop is emotionally charged for real reasons, and no amount of bickering settles it on the spot. What settles it is having picked a method first. There’s no single correct way to divide treasure, but there are several good ones, and the one you agree on before the hoard hits the table matters far more than the hoard itself. Here are six you can put to work tonight, each tested against this exact pile of loot.
Need Before Greed
If you have ever run a dungeon in an MMO, you already know this one in your bones. When an item drops, everyone gets two buttons: roll need if your character will actually equip and use the thing, roll greed if you just want it to sell or stash. Highest roll takes the prize, and every need beats every greed, so the person who can use the item gets priority over the person who only wants the gold it represents. Imported to the tabletop, it is almost embarrassingly simple, which is exactly why it works for the groups it works for.
The rule: When loot drops, each interested player declares need or greed before anyone touches the dice. Anyone with a real use rolls need, anyone who just wants a cut rolls greed, and the highest need roll wins outright. Greed rolls only matter when nobody needed the item, at which point it is a straight contest for who gets to sell it. There is no debate, no negotiation, no waiting for the slowest decision-maker at the table. You declare, you roll, the d20 settles it, and the game keeps moving.
Why it is worth reaching for: This is the method I recommend when a table’s real complaint is speed, when the loot conversation has started eating fifteen minutes of every session and people just want a verdict they can live with. The need-over-greed logic does a surprising amount of work for you: it routes the wizard’s spellbook toward the wizard and the heavy plate toward the fighter without anyone having to argue the case out loud, because declaring need on gear your character cannot use is its own kind of self-exposure. For a group that trusts each other and wants the dice to be the referee rather than the loudest player, it is hard to beat.
Where it cracks, everyone can rationalize need: The whole system rests on an honor code, and an honor code is only as strong as the least honorable person sitting at the table. A +1 longsword drops, and now the fighter declares need because it is his weapon and the cleric declares need because, well, she swings a sword sometimes too and a magic one would be nice, and suddenly the rule that was supposed to end the argument has simply relocated it to the question of what counts as need. There is no objective line. A creative player can construct a plausible use for almost anything, and once one person starts stretching the definition, the incentive to stretch it spreads, because the honest declaration of greed now means voluntarily losing to someone who lied. The honest players eat the loss every time and the shameless ones cash in, and after enough sessions of that, people stop being honest.
Where it cracks, the dice reward the lucky, not the deserving: Even when everyone declares in good faith, the method hands the best items to whoever rolls highest rather than whoever the item should belong to. The rogue who has been begging for that Cloak of Elvenkind for three sessions rolls an 8, the wizard who already owns half the party’s magic items rolls a 19, and the cloak goes to the wizard, prompting the rogue to start arguing that the wizard does not really need it, he just has the highest roll, which is precisely the fight the system was supposed to prevent. The d20 does not know who has had a dry streak or who carried the last fight. It only knows the number, and the number does not care about fairness. Over a campaign, that randomness compounds into the sense that loot is something that happens to you rather than something the party distributes, and that slow accumulation of resentment is the bill this method runs up while looking, session to session, like the cleanest rule at the table.
Round-Robin (Snake Draft)
If need-before-greed’s weakness is that it hands the verdict to the dice, the snake draft takes the dice off the table entirely and hands the choice back to the players. Everyone picks one item, in order, and then the order reverses for the next round, so the person who chose last comes up first when the next round starts. That reversal is the whole trick, and it does more work than it looks like it should.
Run it against the sample hoard, four players deep, and watch the order resolve itself.
Round one, top of the order: The wizard picks first and takes the Cloak of Elvenkind, the single best thing in the pile for someone who wants to vanish mid-fight. The fighter goes next and claims the +1 longsword. The rogue takes the wand of magic missiles. The cleric, picking last, grabs the first potion of healing.
Round two, the snake bites back: The order flips. The cleric, who picked dead last a moment ago, now chooses first, and trading a potion for the wand is an easy call, so the wand changes hands and the cleric walks away with the better prize. The rogue takes a potion, the fighter takes a potion, and the wizard, who opened the whole draft by grabbing the cloak, comes up empty on the back half of the round. The 1,800 gp gets split evenly, because gold divides cleanly no matter what method you use.
Where the snake earns its name: Without the reversal, first pick is simply the best seat at the table forever, and the wizard’s early cloak would be followed by the wizard’s early everything-else. Snaking the order is what keeps the high-value items from pooling at the top. It is not perfect parity, the first pick still gets first crack at the one irreplaceable item, but it is the closest thing to equal agency any quick method offers.
Where it strains: On a dragon’s full hoard with forty line items, taking turns one pick at a time turns a five-minute split into a twenty-minute one. Batch the trivia, draft the standouts, and the snake stays fast.
The Party Fund
The snake draft still ends with four players walking away with four separate piles. The party fund asks a different question than the first two methods. Instead of working out who owns the treasure, you decide what the treasure is for, and the answer is the party. Items go to whoever wields them best for the group, gold pools into a single shared account, and the things you buy with that account are things the whole table benefits from. It is the method that suits cooperative, story-forward groups: the ones who already think of themselves as a unit and would rather sink a windfall into surviving the next dungeon than into four separate piles of coin that never quite do anything.
Run our hoard through it and the assignments make themselves:
- The +1 longsword goes to the fighter, because the fighter is the one swinging in melee every round and the upgrade does the most good in those hands. Nobody else stakes a claim, because the claim was never personal.
- The Cloak of Elvenkind goes to whoever scouts, the rogue most likely, for the same reason: it earns its keep on the body that needs it.
- The three potions of healing and the wand of magic missiles become party stock, carried by whoever has the room and handed out when the moment calls for them rather than hoarded by one player.
- The 1,800 gp is where the method shows its real shape. Rather than splitting into four wallets of 450 apiece, it drops into a shared fund earmarked for shared needs, and your table might agree on the spot that it goes toward a 5,000 gp resurrection diamond, the first deposit on insurance against the day someone does not get back up.
That last move only holds if the table trusts it. You need someone keeping the ledger, and you need contributions to feel even over time, because the fund curdles fast the moment one player suspects they are funding everyone else’s safety net. The question has shifted from who owns it to what serves us, which is the better question; it just hands you a tracking problem in exchange, and that problem is exactly what the table has to solve before the next hoard lands.
Sell Everything, Split the Gold
The party fund leaves the gold pooled and trusts the table to spend it well. If even a shared fund still leaves someone feeling like they got the worse end, this method removes the argument at its root by removing the thing being argued over. You sell the whole hoard, convert it all to coin, and split the total evenly. The +1 longsword, the Cloak of Elvenkind, the three potions of healing, the wand of magic missiles, and the 1,800 gp all become one number, and that number divides cleanly four ways. Nobody claims, nobody negotiates, nobody walks away short. For a table that turns every shiny object into a turf war, that clean arithmetic is worth a great deal.
Why it works: It is the only method that is perfectly equal, down to the copper. Attachment never gets a chance to form, because no item ever lands in anyone’s hands long enough to feel like theirs. Four players, four identical shares, no resentment carried into the next dungeon.
Where it breaks: You are trading away most of what makes magic items fun. Signature gear, attunement choices, the slow build of a character defined by the weapon they carry: all of it dissolves into a spreadsheet cell. The economics hurt too. Magic items rarely fetch their listed value on the open market, and plenty have no buyer at all this side of a major city.
The math, honestly: Say you fence the Cloak of Elvenkind, a 5,000 gp item by the book, and the only buyer in town offers half. That 2,500 gp plus the 1,800 gp in coin, plus whatever the wand and potions clear, gets split four ways, and each share looks thinner than the loot felt when it dropped. Everyone agreed it was fair, and everyone is also a little poorer than the hoard promised.
That coin has to go somewhere, and if your table keeps choosing this method, watch the totals climb. Your players may end up with too much gold and nothing they actually care about owning.
GM Discretion (Story-Based)
Sell-and-split strips every item of its story. This next method does the opposite, because sometimes the loot already knows where it wants to go. The +1 longsword the fighter just pulled from the lich’s vault is the same blade her grandfather lost in the war that opens her backstory, and the moment you describe the notch on the crossguard, nobody at the table is asking whether it should have gone to the rogue instead. That is the whole pitch for letting the GM hand out treasure by story: every reward carries a fingerprint, a reason it landed in this character’s hands and no one else’s.
This method earns its keep when the table cares more about the arc than the audit, because you get to put each reward where it will actually land, in the hands of the player who will feel it. The three potions and the 1,800 gp stop being line items to be divided and become levers you can pull for pacing, holding the gold back as the next plot thread instead of dumping it all in one session.
Where it breaks is favoritism. Do this without saying why, and the player who always seems to walk away with the consolation prize starts keeping a tally, and once that tally exists every future call you make gets read through it, because under this method the bookkeeping is yours and so is the blame for every single decision, the generous ones and the stingy ones alike. The way out is to say the unspoken part out loud. Name the reasoning at session zero: tell the table that you assign loot by story, that the ancestral sword going to the fighter is a feature and not a fix, and that anyone who feels shorted should say so to your face rather than to the spreadsheet in their head. Transparency is what turns a judgment call into a shared one.
The Loot Auction
GM discretion puts the trust in one person’s hands; the auction spreads it across the whole table and lets a price do the deciding. Here is the method everything else has been building toward, the one I keep coming back to because it solves the problem at its root instead of papering over it. Take the party gold and split it evenly, in your head or on paper, so that each of the four players holds an equal claim on it as bidding currency. Then put each magic item up for sale to the table. Players bid against each other, the high bid wins the item, and every bid that wins goes into a pot. When the auction is over, you divide that pot evenly back among all four players. The currency is fictional in the sense that nobody is spending real coin out of pocket; you are just using the shared gold as a measuring stick for how badly each person wants a thing.
What makes it self-balancing is the redistribution. The player who wants an item most ends up paying for it, and that payment does not vanish, it flows back out to everyone, including the buyer. Watch it happen with the Cloak of Elvenkind. The rogue opens at 200, the wizard counters at 300, and because both of them want the stealth advantage badly the bidding climbs in hundred-gold steps until the rogue calls 600 and the wizard, doing the math on what 600 gold is worth to a caster who can already cast invisibility, finally lets it go. The rogue takes the cloak and 600 goes into the pot. Now run the wizard’s books. He lost the item, yes, but when the pot pays out at the end his quarter of that 600 is 150 gold landing back in his account, and the same is true of the cleric and the fighter who never wanted the cloak at all. The rogue, for her part, paid 600 and got 150 of it back, so the cloak truly cost her 450 against the table, which is roughly what it was worth to her and nobody else. The loss softens on its own. Nobody had to argue that the wizard didn’t need it; the wizard decided for himself by refusing to outbid.
That is the elegance, and it is also the cost. This is the most bookkeeping of any method here: bids tracked per item, a running pot, a final division. And it only works if your party actually has a pool of gold to bid with, which means it suits established groups sitting on a hoard far better than first-level adventurers with eleven copper between them. If your table wants fairness without dice and without anyone raising their voice, bookmark this one.
The Six at a Glance
Six methods is a lot to hold in your head, so here is the whole field laid out side by side, because the choice usually comes down to two or three things you already know about your group: how much patience you have for bookkeeping, how fast you want the hoard handled, and how much you trust the table to stay civil when something good drops.
| Method | Speed | Bookkeeping | Conflict risk | Best for |
|---|---|---|---|---|
| Need Before Greed | Fast once everyone knows the rule | Low; a quick verbal claim, no ledger | Moderate; the argument moves to what “need” means | Tables that want a clear default and trust each other to be honest about it |
| Round-Robin | Fast and self-pacing | None; the turn order does the work | Low; nobody can monopolize the good stuff | Groups who hate haggling and just want it over with |
| Party Fund | Slow at payout, fast in the moment | High; somebody tracks a shared pot across sessions | Low on items, higher on how the pot gets spent | Parties saving toward a big shared goal, like a stronghold or a resurrection |
| Sell and Split | Moderate; depends on finding a buyer | Moderate; track coin, not gear | Very low; everything converts to an even number | Mercenary groups who care about gold more than any single trinket |
| GM Discretion | Instant at the table | None for the players; all on you | Low if trusted, high if not | Story-forward tables where loot is a narrative tool and the GM has earned the trust |
| Loot Auction | Slow; bidding takes real time | Moderate; track bids and the redistributed pot | Low; the price itself settles the dispute | Competitive, system-literate groups who enjoy the gamesmanship of bidding |
Tip
Before you pick a method, decide which drops are even worth the ceremony. Trivial consumables can ride along in a batch; the standouts are the ones the table should slow down for. Settle how many magic items a hoard should hold in the first place, and the question of how to split them gets a lot smaller.
Which Should You Use?
If you want one recommendation to start from, default to round-robin or the auction, because both spread the good stuff over time without leaving anyone feeling cheated, and neither requires the table to litigate need on every single drop. The party fund is the right call for groups that already think in “we” instead of “I,” the ones saving toward something collective and happy to let the wallet stay shared. Sell-and-split is the pressure valve, not the everyday method: reach for it only when a particular item is splitting the table, when the fight over who carries it is costing you more than the gold ever could. And let the GM decide when the story is already deciding for you, when a blade so obviously belongs in one character’s hands that putting it up for bid would feel like a betrayal of the fiction. Each of these works. None of them is the answer, because the question doesn’t have one.
What every method shares is the part nobody mentions when they’re arguing about which is best. All of them collapse the moment the table loses track of what was actually found. Round-robin needs an agreed order and a record of whose turn comes next. The auction needs everyone to see the same item and the same bids. The fund needs a running total nobody disputes. Without one shared list the whole table can see, you don’t have a method anymore, you have four people half-remembering different versions of the same hoard, and the loudest version wins by default. That’s the same problem you started with, wearing a different coat.
Picture the four of you back at the dragon’s hoard, except this time you settled the method weeks ago and the list is open on the table. The +1 longsword surfaces, the order says it’s the fighter’s pick, and the sword changes hands in the time it takes to say so. The session keeps moving. The treasure stayed treasure.
Agree on the method once, and the loot keeps moving while everyone is looking at one shared list. Start tracking it free, or bring the question to the table on Discord.
