D&D challenge rating calculator

Hit points put the monster in a defensive band. That band expects a particular armour class, and for every 2 points the monster’s actual AC differs from it, the defensive CR moves one step — a heavily armoured monster with modest hit points ends up rated higher than its hit points alone suggest.

Damage per round does the same on the offensive side, with the attack bonus providing the adjustment. If the monster relies on saving throws rather than attack rolls, use its save DC minus 8 in place of an attack bonus.

The final CR is the average of the two, rounded. That averaging is why a glass cannon and a damage sponge can land on the same rating from opposite directions — and why neither will feel like the other in play.

Damage per round means the average damage the monster deals over the first three rounds of combat, assuming it hits. Not its maximum, not the damage from its best round, and not what it does if every attack lands on the same target.

Recharge abilities — a breath weapon that comes back on a 5 or 6 — should be averaged in across those three rounds rather than counted every round. A dragon that breathes once and bites twice is not dealing breath damage three times.

Multiattack counts in full: if the monster makes two attacks averaging 11 each, that is 22 damage per round, not 11. Forgetting multiattack is the single most common reason homebrew comes out rated far too low.

This calculator implements the Dungeon Master’s Guide method faithfully — and that method does not reproduce the challenge ratings of the monsters in the Monster Manual. Run the published stat blocks through it and most come out around one rating low.

The cause is the defensive half. Published monsters routinely carry fewer hit points than the table expects for their rating, so their defensive CR lands well below their actual one, and averaging drags the final answer down with it. An ogre has 59 hit points, which the table treats as a CR 1/2 monster; the ogre is CR 2.

So use this to compare your homebrew against itself, not to reverse-engineer official monsters. If you have designed something and it comes out at CR 3, a good sanity check is to look at the published CR 4 and CR 5 monsters and ask which yours actually resembles in play.

Even setting the above aside, challenge rating is arithmetic over four numbers. It does not know about resistances and immunities, legendary actions and legendary resistances, at-will teleportation, flight against a party with no ranged options, or a save-or-lose effect that removes a character from the fight on round one.

Action economy is the biggest blind spot. A single CR 8 monster against five characters gets one turn to their five, and will usually lose before it does anything interesting. That is what legendary actions exist to correct, and the calculation does not account for them at all.

Treat the output as a starting point for pricing an encounter, then playtest. If your monster has a mechanic that removes a player’s turn, assume it plays a full rating or two above the number here.