What is a D&D loot table?

A short random table you roll on to decide what a defeated creature, a searched room or an NPC is carrying. Unlike the treasure hoard tables in the core rules, which handle campaign-level wealth, a loot table handles the small scale — the pockets, the pack, the desk drawer.

How much treasure should a monster carry?

Very little, most of the time. Individual creatures carrying coin is fine at a few gold, but the game's wealth budget lives in hoards, not corpses. The value of a loot table is characterisation, not income — a personal object tells the party who they just killed.

Should every monster have loot?

Every monster should have an answer, which is not the same as loot. "Nothing but a chewed bone and three feet of gut string" is a good answer. It rewards the search without inflating the party's wealth and keeps searching worth doing.

How do I make loot interesting without giving out magic items?

Give objects with implication rather than value. An unopened letter from home, a portrait of a stranger, a rock somebody clearly carried for years. These cost the campaign nothing mechanically and change how players feel about the fight they just won.

What dice should a loot table use?

A d12 or d20 per subject is the sweet spot. Twelve results is enough that the same monster feels different twice and few enough that you can scan the page in a second. A d100 table is more variety than you can read at speed mid-session.

Where do loot tables fit alongside the treasure hoard rules?

They sit underneath them. Use the core hoard tables for the dragon's pile and level-appropriate magic item rewards; use loot tables for the five minutes after every ordinary fight, which the core rules do not cover at all.

D&D loot tables: what the monster was actually carrying

TL;DR: "I search the body" gives you four seconds to make the fight mean something, and "fifteen gold pieces" wastes it. Build or buy a table of a dozen results per creature type, weight it towards specific personal objects rather than coin, and keep it to one page you can read at speed. Save the actual wealth for hoards.

The fight ends. Somebody says “I search the body.” And there it is — the four-second window where a session either gains a small piece of texture or quietly loses one.

What usually comes out is a number. Fifteen gold pieces, a shortsword, some rations. All correct, all forgettable, and the scene goes flat before the next one starts. It happens because you are improvising cold, under mild time pressure, on a question you did not prep for, four or five times a session.

That is exactly the shape of problem a random table solves.

What a loot table is for, and what it is not

It is not for wealth. D&D already has a wealth system — hoards, level-appropriate magic items, the treasure hoard tables — and pouring gold out of every corpse breaks it quickly.

A loot table is for characterisation on demand. It answers “who was this creature, in one object?” in the moment the party is looking. The goblin carrying a knotted length of string tells you more about goblins than any stat block does, and costs the campaign nothing.

Get that distinction right and everything else follows. You are not rolling for value. You are rolling for a sentence.

The three-column shape that works

The most useful loot tables have three parts per entry, and the third is the one everybody skips.

The object. A rock. Four feet of string. A locket.
The mechanical detail, if any. Worth 2gp. Worthless. Requires a check to identify.
The implication. He carried it for eleven years. He was proud of the knots. The portrait is of a stranger — he calls her mother.

Column three is the whole product. Without it you have a random object generator; with it you have something that makes a player put down their dice. It also does your improvisation for you: once you have read out that the bandit’s letter from home was never opened, the party will supply the rest of the scene themselves.

Build tables per subject, not one big table

A single universal d100 loot table produces nonsense — a wolf carrying a merchant’s ledger, a lich carrying a fishing hook. Splitting by subject fixes that and makes each table short enough to scan.

Three categories cover almost everything:

Creatures. One table per monster type you use often. Kobolds, goblins, bandits, cultists, wolves, zombies. Twelve results each. You will get more use out of a good kobold table than out of a lich table you will roll once.

People. Townsfolk are searched more often than you would think — a pickpocketed merchant, a bribed guard, a dead innkeeper. Noble, merchant, guard, priest, urchin, sailor, gravedigger. These double as NPC characterisation when the party simply talks to them.

Places. A dungeon chest, a wizard’s desk, an alchemist’s shelf, a kitchen larder, a fresh grave, a tavern lost and found. Location tables are the ones that save you most, because players search furniture constantly and it is never prepped.

150 printable D&D 5e loot tables, one page per monster, townsfolk or location

From the shop

150 Loot Tables for DnD 5e — Printable d12 Book

$34 $10.20

Browse on Etsy →

The MakeMythic book is built on exactly this split — 150 tables at one page each, covering 105 monsters from kobold up to lich, 20 townsfolk and 25 searchable locations, with a d12 per page, a CR index and an ink-friendly edition.

Keep it to a d12 and one page

The temptation is to make tables bigger. Resist it. A d100 table has more variety than you can use, because you cannot scan a hundred rows in the four seconds you actually have. You will end up reading the first entry that catches your eye, which means 90% of the table never fires.

Twelve entries on one page is the working size:

Short enough to read at a glance mid-sentence.
Long enough that the same monster twice in a session does not repeat.
Small enough that you can write your own in ten minutes.

If a result comes up twice, that is fine and arguably good — two goblins carrying the same kind of tat reads as culture, not repetition.

Writing your own

If you would rather build than buy, the process is quick:

Pick a creature you use constantly. Not a dragon. A goblin.
Write twelve objects. Ban coins from at least eight of them.
For each, add one clause about the creature that owned it. Not a story — a clause.
Test it for three sessions and cut whatever never lands.

The hard part is discipline in step 3. “A tin cup” is not an entry. “A tin cup, dented flat on one side, still smelling faintly of the wine somebody else was drinking” is.

For monster stats to go with them, the monster reference covers the numbers side, and monster stat cards keep both in one place at the table.

Running the moment well

Answer instantly. The value of a table is speed. If you are flipping pages, you have lost the benefit — keep the page for tonight’s likely encounters open or printed.

Describe, do not list. Read the object and the implication as a sentence, not as an inventory line. “Fifteen gold and a locket” is a list. “There is a locket around his neck, still warm” is a scene.

Let players ask. After the object, stop talking. Someone will ask a question, and that question is the hook.

Vary the answer to “nothing”. Sometimes the pockets are empty. Say so with detail — that is still a result.

Where it sits in prep

Loot tables are prep you do once and use forever, which puts them in the same category as DM screen inserts and a good note-taking system. If you want the randomised version live at the table, the free loot generator and magic item generator handle the mechanical side while the table handles the character.

Five minutes after every fight, four or five times a session, thirty sessions a year. It is one of the highest-leverage pages you can have open.