I'm worldbuilding an illithid empire. The Lords of Madness book states the dietary needs of an illithid.

A mind flayer must have a minimum of one fresh brain per month. Any less than that and it suffers physical debilitation, becoming so weak that it could die. Its ideal diet is one brain per week. A mind flayer that consumes one brain a week does not feel deprived. It can eat more than that for enjoyment and for the psychic boost, and it will if brains are plentiful.

Given those numbers, what ratio of humans to illithid would be necessary to prevent starvation, and what ratio would be ideal for illithid? Assuming that they're not getting new captives, just breeding existing captives to replace the ones they eat.

I would like the answer to be detailed enough that I could calculate ratios for other races that are capable of reproducing faster or slower than humans, as well.

  • 1
    \$\begingroup\$ You might also be interested in the answers to this question. (Sustainable brain farming and sustainable corpse farming kind of go hand-in-hand.) \$\endgroup\$ Commented Dec 7, 2020 at 14:05

6 Answers 6


It depends, but the absolute minimum is probably 11:1

The answer depends on what assumptions you want to make, though it is safe to say that the minimum to be able to do it reliably is 11 humans to one illithid, with 10 adult females and 1 adult male.

First we assume that the illithid can eat babies. Nothing in your quoted section or in any other material I am familiar with suggests they cannot. We therefore need at least 1 baby per month. If you stagger the pregnancies you should be able to arrange it so that one female from the herd is giving birth per month (Though as Fectin pointed out, perfect timing on that is not needed as long as the pipeline moves fast enough to have at least 1 per month available). That female would then be re-impregnated.

Assuming the illithid used psionics to ensure that the females became pregnant again almost immediately after giving birth and used more psionics to ensure every birth came to term then 9 females can produce one child per month (though delivery earlier should not be a problem as long as it is viable and there are facilities to maintain it until consumed). That is not enough though. You need a pipeline of new females that will actually grow up to replace the ones that wear out. You therefore need to produce a small surplus.

In general, if you are using magic/psionics/etc to ensure prompt impregnation followed by a perfectly timed pregnancy and assuming that a female child would grow to physical maturity at least as fast as the prior one ceased being productive, you would need (gestation period in months females) + (1 extra female to ensure a replacement supply) + (1 male) individuals of a given species to ensure a steady supply that would allow you to have one born per month.

The actual management would be more complicated then that makes it sound because you would need to select which to eat and which to use for the next generation, but the numbers should work.

You might be able to trim it down a little bit by sharing males between illithid herds, but not much. You will need at least enough adult males to ensure that losing one isn't a disaster and while under the assumed ideal conditions one male could certainly handle more than 10 females in this breeding farm way, there are still limits.

Of course, it gets much more complicated if you remove the assumptions about immediate re-impregnation and perfectly timed births. The problem is that its very hard to figure out by how much. We can look up things like average fertility in the real world and average number of miscarriages etc. The problem is that those numbers will likely not apply in this breeding scenario. Average fertility in the real world is influenced by things like how many children couples want to have that would simply have no bearing on an illithid herd. Even average miscarriages vary significantly according to things like healthcare and diet.

So, if the illithid isn't willing to use enough magic/psionics/etc to ensure perfectly timed births and immediate re-impregnation it will need a larger herd, but figuring out how much larger is impossible without know how much effort the illithid will put into encouraging swift breeding and to preserves the pregnancies once they are achieved.

  • 10
    \$\begingroup\$ It's worth mentioning that illithids can eat the brain of an ogre without the ogre dying, and that this results in a literally brainless ogre thrall. This would allow "eaten" ogre slaves to still function as part of the breeding population, which would presumably simplify the logistics of ogre farming compared to, say, gnome farming. \$\endgroup\$
    – GMJoe
    Commented Dec 7, 2020 at 22:23
  • \$\begingroup\$ @GMJoe On what basis are you claiming that a brainless ogre can survive? As far as I'm aware, ogres die from losing their brain just like humans do. \$\endgroup\$ Commented Dec 13, 2020 at 11:20
  • \$\begingroup\$ @EttinaKitten I wish I could tell you. I know I saw a statblock for a brainless ogre thrall somewhere, but I 'm having trouble finding it in my monster manuals... I suspect it may have a non-descriptive name (like "illithid" as opposed to "mind flayer") that I'm blanking on and thus failing to recognise in the indices. \$\endgroup\$
    – GMJoe
    Commented Dec 13, 2020 at 21:35
  • 2
    \$\begingroup\$ @PlayPatrice We are using different assumptions. For one thing, you care about things like recommendations to keep the mothers healthy, I'm assuming our ilithids can handle that through a combination of considering them nearly disposable and psionics. I think your numbers are probably correct or at least reasonable under your assumption set where I'm assuming our illithids who for some reason want the absolute minimum sustainable herd will fix a lot of the problems that you address through caring for the herd and through sheer numbers by using psionics or magic. \$\endgroup\$ Commented Apr 6, 2021 at 23:05
  • 1
    \$\begingroup\$ @PlayPatrice I upvoted your answer incidentally. I think it was detailed and useful, we just approach the issue differently. \$\endgroup\$ Commented Apr 6, 2021 at 23:07

You don't need any humans.

Rings of Sustenance are dirt cheap at 2500 gp. If you've got the resources to build a human farm, then this cost is trivial. In fact, it might be trivial anyway: 2500 gp is below the Wealth By Level for a 3rd level character and even a basic CR 8 Monster Manual Mind Flayer should be well past that point. Even if this character's ring slot is already taken, the Magic Item Compendium gives us rules for adding multiple effects to one item (page 233) and for an effect as cheap as Sustenance, the price will be trivial.

You may argue that this solution is unsatisfying to our hypothetical Illithid. However, the same page that you've quoted has this to say about human farms:

Besides the logistic issues, the brains of lifelong thralls are less satisfying to mind flayers than the brains of free individuals. A thrall has few true experiences to remember and even fewer emotions, which are the "meat and potatoes" of a nourishing, fulfilling mind.

In conclusion, save yourself the effort and just use the ring. Illithids are smart; they would have figured this out. The Monster Manual Mind Flayer has 19 Intelligence!

Ols has pointed out in the comments that the Clear Ioun Stone and the Sustaining Spoon are similar and potentially viable options.

  • 3
    \$\begingroup\$ Disagree with Sustaining Spoon. "contains everything necessary to sustain any herbivorous, omnivorous, or carnivorous creature..." Illithids are not herbivorous, omnivorous, or carnivorous. Their brain-diet is intimately linked to the psychic energy contained in brains, and it the spoon will not provide that energy. \$\endgroup\$
    – Daniel B
    Commented Dec 7, 2020 at 23:53
  • \$\begingroup\$ @DanielB I agree, but I can see some DMs being happy with it. \$\endgroup\$
    – J. Mini
    Commented Dec 8, 2020 at 0:04
  • \$\begingroup\$ The "Clear Ioun Stone" description says Sustains creature without food or water. Arguably Mind Flayers don't require food and water in the sense of the item anyways. The Sustaining Spoon can produce sufficient gruel each day to feed up to four humans. Good for them, useless for Mind Flayers. The Ring of Sustencance though is good. No mention of humans or eating food, instead it refreshes the body and mind which seems ideal. \$\endgroup\$
    – nwp
    Commented Dec 8, 2020 at 13:57

One Brain Per month: 351 livestock per ithilid, including overhead.

  • 96.3 in breeding stock (90 Female, 6 Males, Includes maturity overhead)
  • 238 In maturity overhead for annual / monthly harvesting (Ages 0 to >18)

A wealthy Ithilid Society: 1,068 livestock per Ithilid, including Overhead

  • 293.2 in breeding stock (270 Female, 24Males, Includes maturity overhead)
  • 714 Maturity overhead for annual / weekly / monthly harvesting (Ages 0 to >18)

This answer assumes the following:

  • Rate of Consumption 3.0437 brains per month (Wealthy Societal average)
  • 1.5% Mother Mortality Rate (Medieval Average)
  • 30% Infant Mortality Rate (Medieval Average)
  • 18 Years old is the minimum viable age for both Breeding and Harvesting
  • 24 Month breeding cycle (World Health Organization’s recommendations for humans)
  • An additional 5% overhead added to the numbers to account for general crop losses.

Breakdown and flowchart of logic:

First: How many brains on average do we need to feed an ithilid society?

The minimum is 1 brain per month for minimum subsistence living.

We need to have at least 12 completed pregnancies per year. The world health organization recommends that the beginning of pregnancies be spaced at least 24 months apart for optimal health of the mother and child. So, starting off, we get a minimum value of 24 viable breeding females (12x2).

Now we need to factor in both infant mortality and maternal mortality into the equation. Medieval infant mortality was 30%, and maternal mortality was 1.5% per pregnancy. So let us factor in the 30% expected losses, but still reach the required 12 brains per year.

12 = 70% of X.
12 / 7 *10= 17.1428 - Multiply by 2 to factor in the breeding cycle = 34.2856

then increase the number to account for the expected 1.5% mother mortality rate and we arrive at 34.8077 breeding females to be maintained at all times.

The US Cattle industry recommends no more than 1 bull per 25 to 30 cows under careful breeding conditions to ensure minimum diversity and avoid over taxation of the bulls (which would lower fertility rates). So since we have 34.2856, we will need at least 1.3714 males to cover service our females. Putting our new total up to 35.657. This number represents an absolute minimum active breeding pool necessary just to feed the ithilid. Now we need to factor in replacing our breeding stock as they age out of viable fertility ranges:

Chances of getting pregnant during a 3 month period:

  • 18 percent at age 25 (And I assume below)
  • 16 percent at age 30
  • 12 percent at age 35
  • 7 percent at age 40

The “Chances” of getting pregnant are not relevant for our 24 month breeding cycle, if you begin attempting to breed at month 22 to 26 (5 months), on average you should achieve pregnancy around the 24 month mark. The important part of this table is the viability ranges of fertility. From 18 to 35 we experience our strongest breeding period and once a female reaches 40, they can be safely added to the food supply. So we have an 18 year period in which the breeding females must mature, and a 17 year period in which they are viable. For simplicity sake we will call it 18 and 18. This means the female breeding population must fully replace itself every 18 years.

So an additional 34.2856 females must be born over the course of 18 years, or 1.904 female births per year. Considering there is a roughly 50% chance to get a male or female with every pregnancy, in order to safely breed enough females every time, we will need to double that figure, bringing us up to a total of 3.808 additional live births each year. Factor in the 48 month birthing cycle and mortality rates and we need to add at least 10.88 additional females to our permanent breeding stock to safely reach minimums for both feeding and livestock replacement. Since we should now have a minor overage of males, we do not need to increase our permanent breeding livestock any further to account for their replacement.

So, we are now up to 45.1656 active breeding livestock, and 45.1656 breeding replacements awaiting maturity that we must retain. We also need to add in one additional male and their replacements to maintain healthy breeding ratios. The males can come from the “excess” created by the female breeding, so we don’t need to increase female numbers to account for those. Putting us at a grand total of 96.3312 individuals that cannot be consumed and must be preserved as part of a permanent breeding stock.

Keeping with the theme of maturing livestock – we need at least 12 brains per year for the ithilid to eat – 12x 18 = 216. Roughly 10 “extra” births per year are part of the breeding livestock cycle, and for simplicity of handling, we will simply add them in as extras, bringing us up to 226 livestock in general consumable livestock waiting to mature. Which leaves us with adding in a last dozen of mature “Ready to Eat” livestock to pool, that is assumed replaced by a birth as soon as it is consumed: 238.

So now we have a final “heard” size of 334.3312 humans needed to sustain an ithilid. Let’s go ahead and add in a 5% overhead and margin of safety, accidents and sickness will happen from time to time. Bumping us up to a grand final total of 351.0477 humans per ithilid.

What About Wealthier Ithilid Societies?

For this we will need a wealth distribution chart of some sort. I am assuming wealth is measured by the number of brains you get to eat per month, 1 being impoverished, and more than 4 would be considered extremely wealthy. Note, my numbers are completely Arbitrary:

Impoverished: 1 brain a month (10% of the empire’s population, prisoners and outcasts) Commoner: 2-3 Brains a month (70% of the empire’s population, 2.5 brains for simplicity sake) Wealthy: 4-5 Brains a month (17.5% of the empire’s population, 4.25 brains for simplicity sake) Nobility* : 4-5 brains a week. (2.5% of the empires population, 18 Brains a month)

*The noble classes would definitely abuse their wealth and positions of power in order to maintain their access to nearly daily consumption of brains. This allows them to be more powerful, either to maintain their positions, or protect ithilid society as a whole).

Using this distribution, we come up with a societal average of 3.0437 brains per citizen. Meaning at this point we just have to scale up our previous numbers of 1 brain to 3.x brains to aquire our final minimum livestock numbers:

351.0477 x 3.0437 = 1,068.48 humans per ithilid.

  • 1
    \$\begingroup\$ It should be noted that the Ratio of Cows to humans in the Us. is approximately 3.4 to 1 as a whole. Though gestation periods are similar, the maturity rate is only 1/3 to 1/4 that of adult humans. Beef is not our sole source of dietary nutrition, and one cow producing far more meat than any one person can eat in a single sitting. \$\endgroup\$ Commented Apr 6, 2021 at 1:08
  • 1
    \$\begingroup\$ I think you're missing an opportunity for efficiency: your 10.88 "extra females for breeding" includes doubling the number of births to guarantee females. You really only need 5.44 extras, because your breeding-for-food group will produce plenty of both sexes. But otherwise a thorough approach to a revolting calculation. \$\endgroup\$ Commented Apr 6, 2021 at 6:11
  • \$\begingroup\$ Ahh, the 10.88, instead of the 5.44 extras, that is to completely eliminate the statistical probability of not producing enough females. It's a very small number of extra padding on top of the 5% overhead to ensure that enough food is available to feed myself, and I don't suffer undue crop losses. It's an application of best business practices, cost of event VS cost savings analysis. If we have a female shortfall, we potentially risk falling below our minimum number of viable breeding females, not causing an immediate problem, but a long term one. \$\endgroup\$ Commented Apr 6, 2021 at 6:19
  • \$\begingroup\$ When you're asking for an extra 34.2856 births to replace your breeding stock (second paragraph after the bullet points), you say that you need 1.9 extra females per year. And then you double that 1.9 females per year to get 3.8 extra births per year, because half of your births will be men (who cannot replace the females you're aiming for). But you don't need to care that half of those births will be men -- you have an entire other set of children arriving every year to feed the illithids. So just go back to 1.9 extra births per year, which takes 5.44 extra breeders hanging around. \$\endgroup\$ Commented Apr 6, 2021 at 6:20
  • \$\begingroup\$ ...the point being that you can just take some females from the children-bred-for-brains. You have plenty of females in that group that you can use. \$\endgroup\$ Commented Apr 6, 2021 at 6:23

Using humans is inefficient

Illithids can eat the brains of any humanoid, not just strictly humans brains, so why don't we go with something else? Obviously elves are out of the question, it's generally known that elves have lower birth rates than other humanoids (although whether this is due to biology or culture isn't readily known) Dwarves are in much the same boat, as are gnomes. Halflings and orcs, on the other hand, seem to be roughly similar to humans in regards to birth rates, with halflings having the distinct "advantage" of being small, rather than medium, and so requiring less overall space to ranch effectively.

But goblins, well, those are another story entirely.

Goblins? Yes, goblins! Rather than hundreds of humans, we can get by with only 8 mothers and a single stud per illithid.

But that's such a low number! How can we live off of so few? Well, here are some things we definitely know about goblins:

  • Goblins mature somewhere around 33-55% faster than humans (depending upon edition). They reach maturity anywhere from the age of 8 (1e, 5e) to the age of 12 (3.5e/pf1e, variants in 5e), and live to about 50 years of age.
  • Goblins are small, meaning that they'll require less space to ranch, and less total food over their lifetime.

What's hinted at but doesn't tell us anything specific:

  • According to the entry on goblins from the 4e Monster Manual (pg.136) "Goblins breed quickly and can live most anywhere, from caves to ruins to a city’s sewers"

We can take this to mean they either have a significantly shorter gestation period, or that they reproduce in litters (both commonly accepted theories, and probably both are true). For the purposes of this, we'll assume a little of A and a little of B. Just pulling some probable (and in my opinion, conservative) numbers out of the air, let's say a typical litter size of 2-3 "pups" (avg 2.5), and a gestation period of only 6 months. For reproductive maturity, we'll average the different editions and say 10. This gives us a 33% shorter gestation period, a 150% larger birth pool, and a 45% faster maturation rate.

Other than those explicit differences, we'll assume all other statistics between goblins and humans are essentially equal. Building off of the baseline of using humans in this answer by PlayPatrice, and cutting down on the overall number of needed breeding stock accordingly.

Since we produce an average of 2.5 pups per mother per 6 month period, rather than 1 per mother per 9 month period, we can cut down our minimum stock proportionately. To produce the minimum necessary 12 per year with no sustainability in mind, we need a total of only 3 mothers. Each produces an average of 5 pups per year, leading to a total of 15 per year, so we have a slight surplus. If we take into consideration a health period of 2 years between the conception of each litter for optimal health, the same as a human, then we're only up to 5 mothers total, at a total production of 25 pups per 2 years, leaving us at only a very slight surplus of one pup.

In order to compensate for potential losses (1.5% mother mortality rate, 30% infant mortality rate) we'll of course, need more mothers. Since we can expect about 7.5 infant deaths out of 25 births in every 2 year period, we'll need an increased number of births to compensate, by another 1.5 mothers, which will produce the 7.5 more pups over 2 years. Obviously we can't have half a mother, so that rounds up to 2, for a total of 10 pups. Of course, 30% of those pups are also expected to die, so we'll need one last mother to compensate, bringing us to a total of 8 mothers minimum per illithid.

If we assume that the brains of pups are not nutritious enough due to lack of life experiences, then obviously older specimens are required. If we then say that reaching the age of maturity is good enough for minimal sustenance, then we want to have 12 pups reaching maturity every year, which we can certainly do with our 8 mothers, but the mothers won't last forever. It can be expected that since they die at an age of about half what a humans does, they also lose fertility and stop being viable as breeding stock at a similarly proportioned rate. Since humans have a (roughly) 18 year maturation rate and active fertility period, we'll assume that the goblins are proportionately similar, meaning that with a 10 year maturation rate they'll have a 10 year active fertility period. This means we'll need to replace our breeding stock fully every 10 years, but since we only need 8 mothers, we'll only need to get 8 viable females out of every 10th year. If we assume our goblins mothers aren't all the exact same age, this is fine, and we can replace them easily by selecting the best option out of the current year and exchanging it for the mother that is no longer viable.

All told, we're looking at a minimum of 1 stud, and 8 mothers, with roughly 125 offspring hanging around at any given time, as a permanent sustainable population of goblin brains to eat. You'll occasionally want to go out and capture a "wild" goblin in order to cut down on excessive inbreeding, of course, otherwise mortality rates could go up unnecessarily.

One other note to consider is that goblins are well known for being cannibalistic. If there's not enough food to go around, they'll eat their young. We can turn this in our favor, however, by reducing the overall feeding costs. Since the illithids only eat the brain, and the rest of the body is discarded, we can simply "recycle" it back into the feed, cutting down on the overall upkeep cost. This is a double-edged sword however, as it means that if food supplies are insufficient, then the tasty brains that are meant for the illithids will be killed and eaten before they mature, wasting valuable food resources. It is advised to have the goblins upkeep their own farm with edible plant matter and lesser livestock, in order to sustain themselves without being overly burdening on ones own finances.

As a last note, if we want to get past "the minimum to survive" and move towards "satisfied" then we can simply scale up the total population proportionately by simply multiply the number of total mothers we have hanging around. It will take total of about 32 mothers to produce enough offspring to eat a brain a week, although that means we'll have roughly 500 offspring kicking around at any given time. This is quite a lot for any one illithid to handle, and probably very time consuming to manage, so it's recommended that if one wants this much, they keep the majority of the livestock "free range" and only keep a herd consisting of the minimum necessary size in a more traditional enclosure.

  • \$\begingroup\$ I don't dislike what your trying to do here. ^.^ Author did ask for Human society in their question, which is why mine is tailored for humans. For clarity, you may want to post your final numbers at the top of your post for easy reading / summation as I did. \$\endgroup\$ Commented Apr 6, 2021 at 19:36
  • \$\begingroup\$ @PlayPatrice Yes, this was kind of a frame challenge. I'll add those numbers in. \$\endgroup\$ Commented Apr 6, 2021 at 19:41
  • \$\begingroup\$ Do you also know how much "Maturity" overhead you will need to keep around? You have to have monsters in the pipeline for both feeding at a monthly interval, and replacement of the core breeding slots as they age out. The authors question is asking for ratio of monsters to ithilid - your breeding livestock here, go full ham on it! \$\endgroup\$ Commented Apr 6, 2021 at 19:52
  • \$\begingroup\$ @PlayPatrice Breeding stock and food stock are not separate pools of livestock. When your breeding stock starts to become infertile, you exchange it with one of the freshly matured offspring. Instead of eating the young one, you eat the old one and start breeding the young one. \$\endgroup\$ Commented Apr 6, 2021 at 20:07
  • \$\begingroup\$ The point I am making is that your number is 1 male to 8 females for a total of 9. Given your breeding rate, and rate of maturity, how many additional creatures of non-harvesting age are part of your final total? This affects your ultimate ratio of Ithilid to Livestock. Grand total of 9 assumes you are harvesting the tiny brains of goblins as soon as they are born. - As far as "Inflation" this comes down "Sustainable" farming. The breeding pool takes care of self replacement, the rest is for harvest. The US cattle industry does something similar. \$\endgroup\$ Commented Apr 6, 2021 at 20:09

Mindflayers needs to eat one brain of an intelligent creature per month. But if you eat the brain of a creature it dies. So normally this would expend the creature. But D&D is a world with magic. So when the creature is killed, we could use true resurrection to restore a creature to life even if body parts are missing.

True Resurrection is an expensive spell (25.000 Gp) so lets assume it is used once per day. That way, you have a fresh humanoid each day for an hungry mindflayer.

Besided the cost, you need the approval of the Deity. Good and Neutral Deities will not join because it is cruel to raise somebody with the purpose of killing him again. But Evil deities probably like the idea.

This way you have 30 brains a month enough to supply 30 mindflayers. This amount can be increased if the spell is cast more often.

They could turn this into a movie called Undergroundhog day.

  • 10
    \$\begingroup\$ This answer ignores the fact that true resurrection states "This spell functions like raise dead" which says "the subject’s soul must be free and willing to return. If the subject’s soul is not willing to return, the spell does not work". I doubt most souls would be willing to return just to be eaten over and over again. \$\endgroup\$ Commented Dec 8, 2020 at 17:53

Can they eat Troll brains? Because those would regrow daily... MMMMmmmm. Feasty! (technically in about 15 turns, by HP). Int of 6 counts as intelligent. Provided they take their brains out without killing them (technically possible), their regeneration heals them right up. Just like the Auguary Trolls in Kear Maga (but that's guts, not brains).

  • 1
    \$\begingroup\$ The problem is that the ithilids ability to consume a brain "Kills" outright, and is not a loss of hit points. So the troll's regeneration would not be able to "Save" the troll. In 2nd eddition trolls could go into negative hitpoints, in 3.5 there are no rules for a troll going into negative hitpoints, and it is by convention that trolls have to be burned to prevent ressucitation. \$\endgroup\$ Commented Apr 6, 2021 at 2:28
  • \$\begingroup\$ Yep, that's a problem then. IF they could eat a brain outside a host, they could eat all the troll brains they wanted to. So can they? Or do they only eat brains they pry out with that attack and kill the thing being eaten? Because it seems like it might work if they just "work around" the whole "kill you" thing. Then they could have ioun stones with regeneration working on a body bank supplying brains to eat (outside of hosts). \$\endgroup\$
    – maouse
    Commented Apr 6, 2021 at 2:35
  • 2
    \$\begingroup\$ And could they get a regenerating Ettin, suck out one brain, not kill it, and let it regenerate? \$\endgroup\$
    – maouse
    Commented Apr 6, 2021 at 2:37
  • 1
    \$\begingroup\$ Thats not a bad suggestion, I recomend changing your awnser to reflect that. Make sure to check the rules on what kind of regeneration you need to restore body parts and look over the etin rules before ya post. ^.^ \$\endgroup\$ Commented Apr 6, 2021 at 2:38
  • 1
    \$\begingroup\$ But do trolls die when they are killed? Whether or not sucking a troll's brain out will actually kill it is debatable thanks to the wording of the regeneration rule. (The description of trolls given in the MM does also note that decapitated trolls can regrow new heads, so clearly the original brain can't be too important to them...) \$\endgroup\$
    – Carcer
    Commented Apr 6, 2021 at 7:57

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .