The Deadliest Low-CR Monsters in D&D 5e, Tested in 721,200 Battles
All 133 D&D 5e monsters at CR 1 and below, ranked by how cheaply they break a level-1 party across 721,200 simulated battles. The winner is not the Shadow.
Every "deadliest low-CR monsters" list on the internet is written from vibes: the Shadow's Strength drain, the Stirge's blood-sucking, that one time a Ghoul paralyzed the whole party. We did it the BattleCast way instead - we took every land-capable monster at CR 1 or below in the 2024 SRD, 133 of them, and threw escalating numbers of each at the same level-1 party until it broke, 721K simulated battles in total. The result is a price list for lethality, and the top of it does not look like anyone's listicle.
The deadliest low-CR monster is a pack of hyenas
Not the Shadow. Not the Stirge. A hunting pack of CR 0 hyenas breaks the classic level-1 party for just 165 XP - less than half the 2024 rules' High encounter budget.
The five deadliest monsters per XP
The metric is simple: how much XP does the DM have to put on the field, using only this monster, before the classic Fighter / Cleric / Rogue / Wizard party's win rate drops below 50%? The 2024 DMG calls 400 XP a "High" difficulty encounter for that party. Everything below breaks them for a fraction of it.
Hyena CR 0
Pack Tactics turns every 10 XP body into advantage for the next one. Sixteen of them is a tide of advantage-fueled bites that level-1 AC simply cannot keep out.
Eagle CR 0
Flyers that dive, rake, and force the party to fight on the bird's terms. Half the party has one weak ranged option at level 1, and it shows.
Vulture CR 0
The budget Hyena: Pack Tactics again, plus enough bodies that someone is always flanking. Nineteen of them cost the DM less than two Goblin Warriors' worth of budget.
Giant Crab CR 1/8
AC 15 in a 25 XP shell. Level-1 attack bonuses bounce off it, and every grab pins a hero while the rest of the scuttle closes in.
Giant Weasel CR 1/8
Fast, Small, and numerous. They swarm the back line before the Fighter gets a second swing.
Yes, the camel is in the top ten (~10 of them, 255 XP). Before you laugh, picture ten angry camels bearing down on four first-level adventurers with 8 to 12 hit points each, and remember the 2024 encounter budget claims that fight is barely worth flagging. That is not a bug in the simulation - it is the finding.
XP the DM has to spend before the party's win rate drops below 50%. Lower bars are deadlier per XP. The teal line is the 2024 DMG High budget for four level-1 characters (400 XP) - everything on this chart breaks the party for less.
The 2024 budget prices bodies wrong
The 2014 encounter rules multiplied XP when monsters came in groups, precisely because action economy compounds: every extra body is another attack roll, another flank, another target the party has to spend a turn removing. The 2024 rules dropped that multiplier for simplicity. The gauntlet shows what that costs: at the bottom of the CR scale, the median monster over-delivers, and the effect is beautifully monotonic.
Median XP absorbed before breaking, as a multiple of the High budget, by CR band. A perfect pricing system would put every bar at the teal line. Instead the trend is monotonic: the cheaper the monster, the more fight the DM gets per XP.
The median CR 1/8 monster breaks the party at 0.67x the High budget; the median CR 1 monster needs 1.93x. Put differently: at level 1, a DM buying kobolds gets nearly three times more lethality per XP than a DM buying Ghouls. Pack Tactics is the multiplier the budget forgot - the Hyena, the Vulture, and the Kobold Warrior (#18, ~11.21 copies, 280 XP, 0.7x) all convert crowd size into advantage on every bite.
Every monster that broke the party, positioned by how many High budgets (400 XP) of it the party absorbed before folding. Left of the dashed teal line at 1.0x means deadlier than the encounter math claims. The spread inside each CR row is the real story: CR is an average, not a promise. Hover any dot for the monster's numbers. Monsters past 4x (like the Quasit at 9.65x) are pinned to the right edge.
Where's the Shadow?
The Shadow - the monster every deadly-low-CR listicle puts first - lands at #55 of 108: about 4.82 of them (482 XP, 1.21x budget) to break the party. In an open field where the party can focus fire, incorporeal resistance and Strength drain are not enough to beat the price tag. The listicles are not wrong that a Shadow can delete a lone, cornered Fighter in three rounds - they are wrong that this makes it efficient. Horror and lethality-per-XP are different axes, and only one of them fits in an encounter budget.
The Stirge (#23, ~12 copies, 300 XP, 0.75x) tells the opposite story: mocked as a flying mosquito, it lands at #23 overall even with its blood-drain attachment modeled conservatively. The Ghoul (#97, ~4.56 copies, 912 XP, 2.28x) - the classic level-1 boogeyman - is one of the worst per-XP buys in the band, because paralysis is terrifying but four bodies of it cost as much as an entire kobold warren. And the weakest monster that ever broke the party at all? The Quasit: 19.61 copies and 3923 XP, 9.81x its rating - a familiar, not a fighter.
One more myth for the road: goblins versus kobolds. Per XP, it is not close - the Kobold Warrior (#18, ~11.21 copies, 280 XP, 0.7x) breaks the party 20% cheaper than the Goblin Warrior (#31, ~6.98 copies, 349 XP, 0.87x). We reached the same verdict from a different angle in our goblins vs kobolds tournament: Pack Tactics scales with bodies, and bodies are what low XP buys.
Low-level fights do not degrade - they collapse
The most important DM-facing lesson in the data is the shape of the curve. Against nine hyenas the party wins 98.7% of the time. Against thirteen, 84% - harder, but fine. Against seventeen it wins 45%, and when it loses, it loses everyone: essentially every defeat at that count is a total party kill. At level 1 there is no gentle difficulty dial - the entire distance between "fun pressure" and "roll new characters" is four extra bodies from the cheap end of the Monster Manual.
The party's win rate as the hyena pack grows. Low-level fights do not degrade gracefully - they hold, then fall off a cliff over the span of three or four extra bodies. The dashed line marks the 50% coin flip.
This is why the encounter math failing at the low end matters more than it would at level 10: the level-1 party has no slack to absorb a mispriced encounter. If you take one number from this article to your table, take this one: when the monsters are cheap and numerous, treat the printed budget as a floor, not a ceiling. For tested, hand-built fights at this level, we keep a full difficulty ladder in our level-1 encounters guide.
Replay the breaking points yourself
Every number in this article is reproducible. The embedded battle below loads the hyena pack's breaking-point fight; the links load the classic-monster breaking points into the full simulator, where you can swap the map, the tactics, or the party itself.
The full ranking
Every land-capable SRD monster at CR 1 or below that broke the party within the 24-copy cap, ranked by XP at the breaking point. "Breaks at" is the interpolated group size where the party's win rate crosses 50%.
| # | Monster | CR | XP each | Breaks at | XP to break | x High budget | TPK rate there | Avg rounds |
|---|---|---|---|---|---|---|---|---|
| 1 | Hyena | CR 0 | 10 | ~16.51 | 165 | 0.41x | 56% | 8.8 |
| 2 | Eagle | CR 0 | 10 | ~16.79 | 168 | 0.42x | 52% | 8.8 |
| 3 | Vulture | CR 0 | 10 | ~18.91 | 189 | 0.47x | 51% | 9.6 |
| 4 | Giant Crab | CR 1/8 | 25 | ~8.54 | 214 | 0.53x | 58% | 10.0 |
| 5 | Giant Weasel | CR 1/8 | 25 | ~9.34 | 234 | 0.58x | 62% | 7.8 |
| 6 | Bandit | CR 1/8 | 25 | ~9.64 | 241 | 0.60x | 58% | 8.0 |
| 7 | Guard | CR 1/8 | 25 | ~9.8 | 245 | 0.61x | 53% | 11.0 |
| 8 | Giant Owl | CR 1/4 | 50 | ~5 | 250 | 0.63x | 44% | 8.1 |
| 9 | Giant Rat | CR 1/8 | 25 | ~10.22 | 256 | 0.64x | 61% | 8.0 |
| 10 | Camel | CR 1/8 | 25 | ~10.27 | 257 | 0.64x | 61% | 11.1 |
| 11 | Blood Hawk | CR 1/8 | 25 | ~10.3 | 258 | 0.64x | 62% | 7.4 |
| 12 | Warrior Infantry | CR 1/8 | 25 | ~10.39 | 260 | 0.65x | 60% | 8.7 |
| 13 | Pony | CR 1/8 | 25 | ~10.66 | 266 | 0.67x | 56% | 8.5 |
| 14 | Mule | CR 1/8 | 25 | ~10.71 | 268 | 0.67x | 55% | 8.5 |
| 15 | Noble | CR 1/8 | 25 | ~10.75 | 269 | 0.67x | 54% | 9.6 |
| 16 | Giant Venomous Snake | CR 1/4 | 50 | ~5.38 | 269 | 0.67x | 67% | 5.8 |
| 17 | Pseudodragon | CR 1/4 | 50 | ~5.47 | 273 | 0.68x | 62% | 6.5 |
| 18 | Kobold Warrior | CR 1/8 | 25 | ~11.21 | 280 | 0.70x | 62% | 8.7 |
| 19 | Giant Bat | CR 1/4 | 50 | ~5.63 | 281 | 0.70x | 61% | 9.1 |
| 20 | Flying Sword | CR 1/4 | 50 | ~5.72 | 286 | 0.72x | 58% | 8.6 |
Never broke the party even at 24 copies (25 monsters): Awakened Shrub, Baboon, Badger, Bat, Cat, Commoner, Crab, Deer, Frog, Giant Fire Beetle, Goat, Hawk, Homunculus, Jackal, Lemure, Lizard, Owl, Rat, Raven, Scorpion, Spider, Weasel, Violet Fungus, Dust Mephit, Ice Mephit. Twenty-four cats remain, canonically, not a threat.
How the gauntlet works
The party
The classic level-1 four: Fighter, Cleric, Rogue, Wizard, using BattleCast's 2024-rules class models, fresh resources, and smart tactics on both sides. Same party, same 20x20 open arena, for every single monster.
The breaking point
For each monster we add copies until the party's win rate drops below 50%, then re-run the two straddling group sizes at 2,000 trials each and interpolate the fractional count where the coin flip happens.
The metric
XP at the breaking point: the interpolated count times the monster's XP value. If the 2024 encounter budget were perfect, every monster would break the party at roughly the same XP total. They do not - that spread is the ranking.
What it is not
A combat benchmark, not a table transcript: an open arena with no terrain tricks, no ambush, and no DM adapting mid-fight. A few signature gimmicks are modeled in simplified form - the Stirge's blood-drain attachment, the Imp and Quasit's invisibility games, the Harpy's Luring Song - so read those rows as floors, not ceilings. Fights that stopped making progress (a lone flyer circling forever) resolve by remaining hit points.
Common questions
What is the deadliest low-CR monster in D&D 5e?
Per XP spent, the Hyena. In 721,200 simulated battles, a pack of about 17 hyenas (165 XP) broke a level-1 Fighter, Cleric, Rogue, and Wizard party - less than half the 2024 High encounter budget. Among classic monsters, the Kobold Warrior is the standout: roughly 11 of them (280 XP) do the same job.
Is the Shadow the deadliest low-CR monster?
Not per XP in a straight fight. Shadows broke the party at about 5 copies (482 XP), 1.21x the High budget - slightly weaker than their price tag. Their real horror is Strength drain against isolated characters in darkness, which an open-field benchmark deliberately does not measure.
How many goblins does it take to threaten a level-1 party?
About 7 Goblin Warriors put the classic level-1 party at a coin flip, and the fights are vicious: almost every loss at that count was a total party kill. Kobold Warriors get there cheaper - about 11 kobolds cost 280 XP versus 349 XP for the goblins.
Is CR accurate for low-CR monsters?
It is an average with a huge spread. By XP absorbed at the breaking point, the median CR 1/8 monster over-delivers at 0.67x the High budget while the median CR 1 monster under-delivers at 1.93x. The 2024 rules removed the 2014 multi-monster multiplier, and this is exactly where it shows: numbers beat headline stat blocks.
Related reading
D&D 5e level 1 encounters, tested by difficulty · Goblins vs kobolds: who wins, and on which map? · The best level 1 party composition, tested · The strongest CR 1 monster, decided by 6.8M battles
Want the raw numbers? Download the full gauntlet dataset (JSON) - every monster, every group size, every win rate.