The combat guide ended on a promise: sieges are a separate system, and they deserve their own article. This is that article. Sieges are one of the most confusing parts of EU5 for new players, and the confusion is real, there’s a Steam Community thread literally titled “WTF IS THE SIEGE SYSTEM??” full of genuine, unresolved questions. The replies in that same thread also contain some of the best tactical advice found anywhere in the research for this piece, assembled here into one actual guide for the first time. Combined with the army composition guide, this closes out the full picture of what happens when your armies go to war. You can also watch a siege unfold, dice roll by dice roll, in the interactive EU5 Siege Lab.

Same caveat as the combat article: the wiki’s Combat page, where the siege rules live, is tagged by its editors as potentially outdated, last verified for version 1.0, and the game has moved on through at least a 1.3 patch. The fundamentals here are stable, but treat specific numbers as illustrating how the system behaves rather than guaranteed current values.

How a siege starts

When an army stops in an enemy location that isn’t defended by its own army, a siege begins automatically. Every location has a Garrison number, and the siege is won once that garrison is either killed or its morale collapses entirely.

The siege just runs once you sit there. You don’t need to issue a special order, you need to be present in force, and the garrison number tells you what you’re up against.

The minimum force requirement

To make progress at all, you need at least as many soldiers besieging as the location has in its garrison. For an unfortified location, at least 100 men will occupy it in 15 days, though various factors can change that timeline.

This is why the army composition guide keeps hammering on force size. If you show up to a siege under-strength, you don’t slowly chip away, you simply don’t progress until you bring in more men.

Assault, the fast but bloody option

The attacker can order an assault at any point. It speeds up siege progress dramatically, but it costs soldiers, and once ordered, it cannot be cancelled. Up to three Breaches can occur during a siege, and each one reduces how many soldiers die in an assault.

That last part matters more than it looks. A siege that’s gone on longer and produced breaches makes a later assault safer than an early one would have been. Rushing an assault before any breaches exist is the bloodiest possible way to take a fort.

Phases and the dice roll table

Each siege phase lasts 30 days by default, adjustable by various factors. Every phase rolls a die, and the result drives what happens. Fort Level reduces that roll, so harder forts resist longer. Each Breach increases it by 2, and artillery can add a further bonus. Here’s what the roll actually produces, in plain language:

• A roll of 1 triggers a Disease Outbreak.
• Rolls 2 through 4 mean nothing changes, the status quo holds.
• Rolls 5 through 11 cause a Supplies Shortage for the defender.
• Rolls 12 through 13 cause a Food Shortage.
• Rolls 14 through 15 cause a Water Shortage.
• Rolls 16 through 19 cause defenders to desert.
• A roll of 20 means outright Surrender.

Notice what this table is really saying. Progress toward a win is a climb, and most rolls land somewhere in the middle, chipping away at the defender through shortages before desertion and surrender become possible. Sieges in EU5 are a lottery ticket you can tilt, not a button you press.

The blockade detail that actually matters

A coastal location that isn’t blockaded gives the attacker a -1 penalty to the dice roll each phase. That’s a specific, real number, and it directly confirms the advice you’ll see in every serious siege discussion: blockading a coastal fort isn’t just generally helpful, it’s removing a penalty you’d otherwise eat every single phase.

If your target is on the coast and you can get ships there, blockading should be near the top of your to-do list. It’s the difference between rolling normally and rolling at a permanent disadvantage.

What artillery contributes

Artillery helps, but how much depends on its condition. Each artillery regiment contributes to the siege dice roll based on its strength and the fort’s level, a full-strength artillery unit contributes meaningfully, a heavily depleted one contributes very little or nothing at all.

The plain-language version: stronger, fuller artillery units meaningfully speed up a siege, while a badly depleted artillery regiment barely helps. Dragging a half-strength artillery stack to a siege is a lot of travel for almost no gain, one more reason to keep your artillery healthy between battles.

Tactics the community has actually tested

This next batch comes from player experience, not the wiki, and it’s the best stuff in this guide. Treat it as community-tested knowledge rather than officially documented developer statements.

The AI doesn’t pay to keep forts fully staffed during peacetime. Garrisons are typically only around 33% strength the moment a war begins, which is a shock to anyone assuming they need to prepare for full-strength defense.

The timing of your declaration matters. Declaring war on the first day of a month reportedly gives roughly 30 days where assaulting forts is close to free, because the garrisons are at their lowest and haven’t ticked up yet.

Regulars are strong at assaulting. A full stack of regular troops can reportedly take 3 to 5 forts in the first month of a war using those assault bonuses, before the garrisons recover.

Generals matter here too. A general’s Diplomacy and Admin skills can boost siege effectiveness, in some cases up to 100%. This connects straight back to the combat article’s finding about Diplomacy affecting morale, the pattern is consistent, generals are mechanically more important across multiple systems than their flavor text suggests.

Ships can participate directly. From a certain ship tech level onward, naval units in adjacent sea tiles can join battles and sieges directly, and blockading a coastal fort can massively cut down siege time.

And starving the defender out is a legitimate strategy. Buying up or controlling a besieged nation’s food supply is a real, usable tactic, not just flavor text, and it compounds with the Food Shortage rolls on the table above.

What’s different from EU4

If you’re coming from EU4, a few habits will actively mislead you here, and these differences trace back to the game’s lead designer, per reliable player discussion. There is no instant create a wall breach button the way EU4 had, that shortcut is gone. Assaulting burns through an army’s morale much faster than it did in EU4, so it can’t be spammed the same way you could there. Food logistics are now a real constraint during a siege, a besieging army that isn’t fed can starve and desert, which never worked that way in EU4.

Most importantly, the system is intentionally built to be uncertain. The design intent, per the lead designer, is that sieges should be RNG-driven on purpose, both to simulate the psychological uncertainty of a real siege and to make it genuinely possible for the besieger to fail outright.

That last point deserves to be said plainly to any EU4 veteran: don’t expect siege outcomes to be as predictable as they were in the previous game. That unpredictability is a deliberate design choice, not a bug. And honestly, it’s a choice that will either frustrate or delight you depending on how much you liked the old system’s predictability.

What this means for your war plans

The full picture across three articles now runs together. The army composition guide builds the force, the combat guide explains what happens when that force fights in the field, and this guide covers the sieges that follow. Bring enough men to match the garrison, keep a healthy artillery stack, blockade coastal targets, and time your assaults to the first days of the war. Whatever the dice roll, you’ve given yourself the best chance the system allows, and for the forts you really can’t afford to roll poorly on, you can starve the garrison out before the lottery even starts.

Try It Yourself

Want to feel the lottery before you bet a war on it? Our EU5 Siege Lab lets you pick a fort, a garrison, and your army, then rolls the 30-day dice phase by phase, showing you the modifier math and an autopsy of exactly why the fort fell or held. It’s the fastest way to internalize why artillery, breaches, and blockades matter more than raw siege length.