The castle is under siege. The inhabitants are running out of food but have plenty of gold.

The attackers have food supplies but need money to keep them coming.

Does it make sense for the attackers to sell food to the castle-dwellers at inflated prices in order to get all their gold rather than simply starving them out at the beginning?

If this is so then why has no-one ever done it?

I think perhaps I should say something about why I think it's a good idea.

If you simply starve the castle-dwellers from the start then they will likely bury all their valuables in the hope of recovering them later. They will also start eating valuable animals such as horses.

If you give them food in exchange for their valuables of all kinds then you have both the gold and the valuables. Once they no longer have anything worthwhile to pay you with, then you can start to starve them out.

  • 24
    $\begingroup$ What happens to the gold if the siege is successful? $\endgroup$ – a CVn Dec 25 '18 at 19:02
  • 11
    $\begingroup$ From a simple accounting perspective, this makes no sense: Sieges are massively expensive. The probable cost of paying, equipping, and feeding the enormous besieging force each day is likely to be much greater than the wealth you can extract from the besieged enemy that day. There are much cheaper and easier ways to acquire access to the hoarded wealth: Blackmail. embezzlement, marriage, alliance, etc. Shakespeare would write a devious protagonist who pretended to ally with the wealthy castle-owner, spending the castle's money without even needing to move it out of it's current location. $\endgroup$ – user535733 Dec 25 '18 at 20:15
  • 41
    $\begingroup$ Why would the defenders trust the attackers not to add poison? $\endgroup$ – Chronocidal Dec 25 '18 at 20:51
  • 15
    $\begingroup$ Historically there have been some cases similar to the one you are envisioning, but it didn't end they way you think. If the besieging army is only looking for plunder (they're there for the gold, not trying to grab land or any other purpose) they will usually accept a (large) payment in gold to pick up their belongings and go away. Constantinople used to do that to several armies who went to camp outside its walls. It was just cheaper than waging war against them - and you couldn't loose. $\endgroup$ – Rekesoft Dec 26 '18 at 8:12
  • 5
    $\begingroup$ Might depend on who’s doing the selling. As in I, personally, would happily part with some vegetables and commit treason in exchange for enough gold to bribe my way out of military service, buy a beach frint hut somewhere warm and retire before my my death of old age at 35. $\endgroup$ – HopelessN00b Dec 26 '18 at 10:27

16 Answers 16


It makes no sense, as the attackers will obtain all the gold of the besieged anyway, as soon as the defense has fallen. Selling food to the defenders is the worst possible action, as it increases the losses of the attackers who need to resupply, keep their troops healthy and motivated without being able to do much. Time is essential in a Siege. The longer it takes, the higher the risk of winter, reinforcements, changing alignments breaking the siege. And hunger is your best weapon, breaking your enemies' resolve, morale and ability to fight most reliably and within a reasonable amount of time.

  • 2
    $\begingroup$ I think it makes some sense. As I said above, "The point is that the besieged might bury or hide their gold if they are simply starved out. They might hope to survive and dig it up later. If they are forced to pay for food then the attackers get all the gold before starvation starts." $\endgroup$ – chasly from UK Dec 25 '18 at 19:20
  • 31
    $\begingroup$ The winners will have plenty of time to discover the gold's hiding places. They can always "gently encourage" a few survivors to remember. $\endgroup$ – user535733 Dec 25 '18 at 20:06
  • 1
    $\begingroup$ "the attackers will obtain all the gold of the besieged anyway, as soon as the defense has fallen" - but what if the attackers need the gold to pay their mercenaries, without whom they won't be able to take the castle? $\endgroup$ – vsz Dec 25 '18 at 21:04
  • 6
    $\begingroup$ @vsz - if they need the gold to pay their army, it behooves the besieged to wait them out. $\endgroup$ – jdunlop Dec 25 '18 at 23:30
  • 5
    $\begingroup$ It's only mentioned in passing but the threat of reinforcements is the big one: the longer the siege takes, the more likely it is that the defenders can round up a bigger army and march it over to break the siege. That's the ultimate point of fortifications after all - make the siege take long enough that you have time to send over your army and drive off whoever's trying to attack you $\endgroup$ – Pingcode Dec 26 '18 at 0:52

In tribal warfare trading with the enemy is normal, even in the modern age in conflicts like the ongoing Syrian war. The factions were trading fuel and weapons, even though those same weapons would likely be turned against them. So the concept itself is not entirely irrational. It's very much a matter of the sort of war you're having.

However, half the purpose of a siege is to starve the castle out, so you're not going to be trading them food. Once the castle runs out of food you've won, so you're not going to act in a way that extends the siege beyond the minimum necessary, regardless of any other potential returns.

If you're worried about them burying the gold, simply force them to ransom themselves. They have to give you the gold in exchange for free passage out.


This would be a good thing for a story.

In general, those in charge who are directing a siege want to starve out the occupants of the city under siege. This is because the besieged site for whatever reason cannot be taken by direct military action.

The troops who are actually conducting the siege might have very different motivations. These commoners might have been drafted to fight, compelled to stay by force or stay in the hope of getting loot when the city falls. Individuals among these common soldiers might very well surreptitiously take the chance of selling food to persons in the city under siege. These individuals make some profit in the short term - which makes sense if maybe the city is not going to fall or for some reason they are not going to get any loot. If someone catches one of these opportunists, that person might be in trouble - or might just have to hand over half the gold...

Blockade runners could be considered an example of this. A naval blockade is a type of siege. A port city dependent on resupply by sea is prevented from resupply by the blockading naval force. The blockade runners evade the blockade to resupply to city. Could the blockade runners be members of the same navy that is doing the blockade? A wily captain might try to butter his bread on both sides this way.


It's a genius idea, if you're the attacker.

Poison the food, sell the food. Get the gold and the castle after a few days only.

In all seriousness, if you're undersiege you should be very wary of anything the attacker would give you. I don't think accepting food from it is very smart.

And from the attacker point of view, giving food without poisoning it will prolong the siege and it isn't something you want.

  • If your army is here, it isn't somewhere else defending your towns and forts which are open for an attack from another army.

  • The defenders may be waiting on reinforcements to break the siege.

  • 1
    $\begingroup$ Not sure if "genius" is misspelled on purpose or not. $\endgroup$ – jpmc26 Dec 27 '18 at 0:31
  • 1
    $\begingroup$ Mh, probably not. $\endgroup$ – Echox Dec 27 '18 at 8:17

Negotiate their surrender

If the besieged are starting to feel so hungry that they are ready to trade valuables for food, it's a bit like a partial surrender. Before that, their point was that they could protect their valuable from you. Now they need your goodwill to keep them safe. Then you know it's a good time to start negotiating their full surrender instead.

After all, surrender is this: give us what we want and we'll give you what we want in return. So you can offer:

Give us your gold, horses, access to your castle, head of your king why not, etc. And we will give you some food, you stay alive and we reduce the military presence at your door a bit. From now on you pay taxes to us and your military serves us or die.

Instead of only offering some food for some gold.


This only makes sense if the gold is what the attackers want

If the attackers are looking for land or political influence, then this doesn't make sense at all. Why prolong the engagement if the gold isn't the #1 reason you're there? Why risk your people and your own supplies?

If the gold is what you want, selling food until you have all the gold is a dandy solution. If the gold isn't the single most important reason you're there, this is a remarkably foolish tactic.

  • $\begingroup$ Yes, but now you have all the gold and then you can stop supplying food and start starving them out. Or indeed you can ask for other valuables. $\endgroup$ – chasly from UK Dec 25 '18 at 19:21
  • 3
    $\begingroup$ @Chasly if you were going to untimately take the castle why not just start off doing that rather than delaying it. This only makes sense as a form of extortion; where you get most of the rewards without having to risk getting killed storming the walls $\endgroup$ – Richard Tingle Dec 25 '18 at 22:52
  • 4
    $\begingroup$ My point is, if the gold isn't the only reason you're there, then ultimately you pay a higher price by giving them time. sieges are never simple. The cost of posting & supplying the army. The unreast caused by their absence. The loss of strength and prestige while they're away. Better hope nobody lays siege to you while you're wasting time bartering for some gold. $\endgroup$ – JBH Dec 25 '18 at 23:11

The 'gold' would need to be something the defenders can easily get but the attackers cannot.

Whether it's elves and mana, molepeople and minerals, or whatever, it needs to be something that the attackers wouldn't be able to (easily) get once they kill all the defenders. It costs a lot more to enact a siege than to be the one being besieged, and the main advantage is that resources aren't coming in to the besieged. So for them to draw out the siege (by feeding the defenders), it would have to be something worth a lot to the attackers, enough to be willing to draw out an expensive siege, and something they can't access themselves once inside the walls.


No besieged castle will ever buy food off the besiegers for the fear of poison or disease no matter the cost.

All it takes is a tainted batch of food and suddenly you start to lose your defending force and weakening the rest.


This is Normal

First of all, this is the case in so many other contexts as well. "Your money or your life" If the armed robber is really ready to kill someone, they won't even ask the question. They do ask the question because they don't really want to kill someone, they just want the money. But they are ready to kill if that's the only way to get the money.

Similarly, the attackers don't really want to kill people, or have them die of starvation. They'd rather get the money, the castle and slaves. So they are perfectly content to get the gold and wait on the castle. The people willing to trade rather than starve are the same people who will, in the end, rather surrender and become slaves than die defending the castle. Well, the leaders will probably get killed anyway, as leaders tend not to make great slaves, but most of the people in the castle were effectively slaves to the owners of the castle anyway, so they'll gladly become slaves of the other side instead of dying of starvation if that is an option.

Where do the attackers get the food? From the peasants surrounding the castle who have already been beaten into submission. So they don't actually spend the gold on food, they spend it on luxuries and send part of it (reportedly "all") home to their king.

For those who say that they would never do such a thing, keep in mind that a common way of managing far-flung lands was to have them pay tribute (a.k.a. exorbitant taxes), so getting a steady income stream from the castle was really just a prelude to the long-term situation. (Until the revolt...)

  • 4
    $\begingroup$ But is it normal in a siege? Do you have any evidence for this? $\endgroup$ – chasly from UK Dec 25 '18 at 19:06
  • $\begingroup$ @chaslyfromUK I don't know. But this is WorldBuilding, not History. $\endgroup$ – manassehkatz Dec 25 '18 at 19:10
  • 1
    $\begingroup$ @chaslyfromUK In the era of castle building in medieval Europe it was not common to take slaves from capturing a castle, slavery had been replaced by serfdom in many places. There were some slaves particularly of other religions such as captured Muslims in parts of Southern Europe but it was not widespread. $\endgroup$ – Sarriesfan Dec 26 '18 at 7:16
  • $\begingroup$ You say its normal, but I am unaware of anytime in the history of warfare this was done in context of a siege. Also, the "peasants surrounding the castle" normally come inside the walls before a siege. If they anticipate a siege, they will likely harvest everything harvestable and burn the fields on their way inside the walls... $\endgroup$ – TimothyAWiseman Dec 28 '18 at 23:11
  • $\begingroup$ @Sarriesfan While you are right in a sense, it is worth noting that many historians consider serfdom to be a form of slavery. $\endgroup$ – TimothyAWiseman Dec 29 '18 at 0:05

What would be much more likely to happen in real life would be that the city would give the attackers the gold in return for their going away.

Historynet.com says that

The most satisfying way to successfully conclude a battle was without fighting. Indeed, many more sieges were settled by negotiation, bribery, or forms of intimidation than open warfare.

This works if the attackers are in it mainly for the money. Once they're paid, they've gotten what they wanted, and the city does not have much gold left, so continuing the siege for the remaining gold would cost more money than could possibly be gained.

And if the attackers are not in it just for the money, then if they feed the besieged city, it is going to take much longer for them to surrender. And once the city is conquered, they are very likely to get the gold anyway.


This seems like something that could happen under very specific circumstances, but it's not particularly likely. It would mean that both the attacker and the defender planned their actions terribly. In the case of the attacker they embarked on a siege attempt without making sure they had enough money available to complete it and had no more-reasonable contingency plans. In the case of the defender they had a big pile of money that they should have spent on larger storehouses and more provisions before anyone came to besiege the city.

However, if both of those miscalculations happened, then... maybe? For example, if the attacker is short on funds to make payroll and morale is low enough that the only options are sell food to the city for gold to pay their troops or go home, the attacker might well make the offer. If the city is nearly starving but knows that reinforcements will be arriving shortly to break the siege if they can just hold out a little longer they might accept such an offer. Of course, the attacker would have to not know about the reinforcements or they'd be likely to just pack up and leave, and the city would be highly suspicious that the offer was either some kind of trick, or evidence that the attacker was about to run out of money and would be gone if they just hold out a bit longer... Diplomacy is complicated.

Of course, if the attacker is solely in it for the money they will probably just demand some form of ransom payment ("danegeld") in exchange for lifting the siege and going away. If they have a good reputation for actually going away when paid the city would be quite likely to take them up on such an offer if it weren't too expensive.


Generally, this doesn't make sense. However, there are a few cases in which this could work:

The most plausible example comes from history, where castles were sieged for gold. The castle would sometimes pay a massive sum for the siege to end. This is, in a sense, buying food.

Another example would involve a plan to poison the food and then sell it to the castle. Obviously, the defenders would be stupid to buy poisoned food. However, such an attack could be massively more successful if the attackers first sold a smaller portion of non-poisoned food to build up trust or if the poison was slow acting.

In a final case, if the attacking army has individuals who are greedy and not loyal, those individuals could secretly obtain and sell some food to the defenders in exchange for gold, which they would personally keep and profit from. Obviously, the certain penalty if they were caught would be death.

I can't think of any other ways this would work.


It really depends on the reason for the seige. "starve them out" isn't the only reason to seige. Another is resource denial, e.g. if the result of the seige is that you cut the supply line of a third party, e.g. your enemy in a larger war, or another kingdom you want to weaken. Or keeping the denizens of that city from traveling to another place to petition a king, do an election or census.

It also may be impracticable for the enemy to pinch off the food supply due to their inability to control smuggling.

There is plenty of precedent for humanitarian aid to the enemy, ranging from the Japanese "give salt to the enemy" to Hoover's food aid for the Communists to Operation Chowhound during World War II/Great Patriotic War.


No one has ever done it because it would be self-defeating and probably unnecessary.

For the defenders, a castle could have a year's supply of food and its own well.

On the attacking side, a castle siege wasn't something you would enter into lightly. It took a great deal of preparation and planning and time was typically on the side of the defenders for several reasons:

  • A medieval army didn't have a supply line as a modern army would. They would raid the local area for supplies. After a time they might not actually have that much food themselves, especially if the loyal local peasants have implemented a scorched earth policy, killed all the livestock, burned the fields and run away.
  • Medieval armies take time to put together and to travel anywhere. The longer the defenders can hold out the more time their side has to send more and better help.
  • Noblemen typically would be held to give their king/next person up in the hierarchy a certain number of fighting days each year. Once those days are up, they'll want to go home and they probably will. As a defender, if you can hold out long enough the besieging army might just dissipate before your eyes.
  • A medieval army is very vulnerable to disease. Imagine a mobile village with very limited sanitary facilities and no meaningful medicine. Dysentery was rife and anything infectious could and did rip armies apart. The larger the force and the longer it sits in one place, the more vulnerable it is.

The defenders don't want to die and the attackers don't want to die. So, in practice, what would often happen is that the siege would be a negotiated affair. The besiegers would offer a deal like, "if you're not relieved in 30 days then give us the castle and you can go free". Honour would be satisfied on both sides (with no repercussions) and each would have a way out without everyone getting killed. That would also give the defenders time to contact their superiors and find out what they want them to do.

So there you are. You're a medieval commander, you know you're on a clock, and the castle defenders send out a messenger to ask you to trade for food so they can hold out for a bit longer. Do you trade or do you get a few of your less healthy looking men to defecate into the largest pot they can fill and catapult it into the middle of the castle along with the severed head of the fool messenger they sent out?


"The enemy" isn't a single body with a single mind. There are power-hungry generals, corrupt captains, opportunistic corporals, etc. All occupying military powers I can think of, had issues with black market deals. My family was saved because someone paid a corrupt Nazi general with creates of gold for their lives.

Your food-for-gold deal doesn't have to be negotiated by the leaders of the opposing sides. The deal can be struck between a shady banker (or a mobster, a thief, a benevolent merchant, whatever), and a corrupt general (or supplies officer, or a driver, etc.)

Self interest always wins.


Tactically it's a very bad idea. As mentioned in other answers, it only make the Siege harder. However, under certain strategic and political situation, this can be a plausible action.

Let's first analyze the possible benefit of doing so for the siege side (considering the side under siege would accept the offer):

  1. Possibly keep the army under siege where they are, so there won't be any unexpected activity (like trying to break the siege out of desperation)
  2. Reducing the casualty in the castle, this can work in all different ways that I will elaborate below
  3. Getting the gold right away

So yes, as long as breaking into the fortress isn't your primary goal, we can imagine some scenarios that make this plausible:

A token force trying to keep a fortress and its army in a strategically non-important place contained without paying too much casualty.

Bonus if the local vassal also don't want to get involved in the larger scaled war but still have to show some level of resistance, then a hidden deal can be signed, the siege side maintain the siege, while also keep the fortress army well fed so they won't do anything unexpected.

A war that near ending, yet there might be a key general under siege in a fortress, which can be a huge interruption for any future war effort for the attacking side. Thus attacker can lure the defender with a offer: provide the fortress with needed food for money.

Trading with enemy during war time is a treasonous act, and can provide some serious political tempest inside the defenders, which might be even more efficient than just killing the general. A general killed in war is a hero; a general getting executed after war for politically reason though, with the generals supporters, can easily split or even destroy a country from inside.

Or of course, you can always keep the truth as a timed bomb and keep the general's hand tied, or even better, create a puppet in the enemy military command system.

Again a war is ending, and the fortress under siege is part of the truce deal, yes, the attacker side will be their new lord. Thus attacker may choose a softer act to avoid frictions in future rules, after all, it's very likely that the soldiers inside the fortress are just drafted from villages around it.

The Great Khan is dead, and we are all going back home for the new Khan's election in two weeks. It's very unlikely we can force ourselves into the fortress in two weeks, not mentioning the casualty we have to take (soldiers are important resources considering the upcoming battle for the new Khan's throne). How about let's make a deal with the guys in the fortress for some gold that may be useful in future.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.