This question: Can you get enough water by eating only fish? asks if a person could survive on fish alone. Can a person survive on fish and/ or blood alone of any species if stuck at sea or animal blood as a last resort where there is no water or fire?

Obviously if it was a fresh water fish there is water, but there are fresh water mud skippers that can breathe air and the water to tainted to drink in that case a fresh water fish blood maybe safer than the water. https://chemistry.stackexchange.com/questions/115901/what-can-i-eat-that-will-help-metabolize-blood

Desalination would be the best way to process the blood but this is in emergency situation scenario.

From @PTwr Comment's Link: If you drink blood regularly, over a long period of time the buildup of iron in your system can cause iron overload. This syndrome, which sometimes affects people who have repeated blood transfusions, is one of the few conditions for which the correct treatment is bloodletting. https://what-if.xkcd.com/98/

  • 2
    $\begingroup$ What level of processing is allowed? $\endgroup$
    – Peter
    Commented May 22, 2019 at 13:56
  • 8
    $\begingroup$ Obligatory XKCD reference. $\endgroup$
    – PTwr
    Commented May 24, 2019 at 11:23
  • $\begingroup$ @PTwr good find . I have included some of it into the question. Thanks $\endgroup$
    – Muze
    Commented May 24, 2019 at 16:51
  • 1
    $\begingroup$ Note fish blood contains a lot less hemoglobin. some fish lack it entirely. $\endgroup$
    – John
    Commented May 25, 2019 at 3:57

4 Answers 4


Blood is not a good source of water.

1 liter of blood contains about 800 mL of water, 170 grams of protein and 2 grams of sodium (calculated from the composition of lamb blood).

When metabolized, 170 grams of protein yields the amount of urea that requires 1,360 mL of water to be excreted in urine (calculated from here); 2 grams of sodium requires about 140 mL of water to be excreted (from here).

This means that drinking 1 liter of blood, which contains 800 mL of water, will result in 1,500 mL of water loss through the kidneys, which will leave you with 700 mL of negative water balance.

Fish blood can contain less protein, for example, trout (check Table 1) contains about 120 g of protein (plasma protein + hemoglobin) per liter of blood. Using the same calculation as above (1 g protein results in the excretion of 8 mL of urine), drinking of 1 liter of trout blood, which contains about 880 mL of water, will result in 960 mL of urine, so in 80 mL of negative water balance.

Turtle blood can contain about 80 g of protein (plasma protein + hemoglobin) and 3.4 g of sodium per liter. Drinking 1 liter of turtle blood, which contains about 920 mL of water, will result in 80 x 8 mL = 640 mL loss of urine due to protein, and ~240 mL due to sodium, which is 880 mL of urine in total. This leaves you with 40 mL of positive water balance (to get 2 liters of water per day you would need to drink 50 liters of turtle blood, which isn't realistic.

In various stories (The Atlantic, The Diplomat, The Telegraph), according to which people have survived by drinking turtle blood, they have also drunk rainwater, so we can't conclude it was turtle blood that helped them. I'm not aware of any story that would provide a convincing evidence that the blood of turtle or any other animal is hydrating.

  • 2
    $\begingroup$ I think this answer addresses the main reasons. The iron issue may be a long term problem but most single foods will cause long term problems if one tries to feed on them exclusivley. $\endgroup$ Commented May 22, 2019 at 14:22
  • $\begingroup$ Assuming a sufficient supply, could you drink a lot of blood and eventually get to a point where the protein is no longer metabolized? $\endgroup$
    – Flater
    Commented May 23, 2019 at 11:19
  • $\begingroup$ Normally, protein does not accumulate in your body, so it is metabolized even if you consume it in excess. Excessive protein is mainly converted to body fat. $\endgroup$
    – Jan
    Commented May 23, 2019 at 12:18
  • $\begingroup$ Wouldn't a dehydrated body stop even picking up the protein from the small intestine and only absorb water? (just a guess) This also doesn't get rid of the sodium problem, but that is not so severe.. $\endgroup$ Commented May 23, 2019 at 13:50
  • $\begingroup$ Protein absorption would stop only in severe dehydration. As long the water is absorbed, the protein will likely be absorbed too. $\endgroup$
    – Jan
    Commented May 23, 2019 at 13:53

You can drink blood of course to a minimalistic amount (eg- a few teaspoons ) and also if blood is free from pathogens. But it should always be in very small amounts and from suitable donor. Here's why

The strange fact is, blood, when drank, is toxic. When confined to places where blood is supposed to be — such as the heart, vessels, and so on — it is essential for life. But when ingested it's a very different story.

Now why is it even harmful?

  1. Haemochromatosis- this is a disease where your body has excess iron/ iron overdose which gets deposited in organs such as heart, liver , pancreas etc. This happens because blood is very much rich in iron but our body doesn't have any suitable mechanism to excrete the excess iron. (There is some doubt with the actual cause of Haemochromatosis. It is a disease related to iron overdose but its not sure whether the main cause is genetic or it can be anything related to diet as well.)

  2. Blood related disorders, eg :- AIDS, Hepatitis B etc - If blood is not taken from a suitable donor or the equipments you use to take it in are not sterilised properly, you'll probably contract one of these disorders, which are absolutely life threatening.

  3. Malnutrition- Although blood contains 93% proteins and 1% carbs , blood is terribly low in minerals and vitamins.

So basically blood taken in small amounts will have no harm on you, but you surely can't survive on blood.

For the sources, visit - https://www.livescience.com/15899-drinking-blood-safe.html



  • 4
    $\begingroup$ @Arsak thankyou so much for the suggestion, I have added the sources. $\endgroup$
    – Ishi
    Commented May 22, 2019 at 6:03
  • 3
    $\begingroup$ Haemochromatosis is a genetic disorder found in those of European descent, and common among those with Irish descent. It has nothing to do with drinking blood, but involves a variant chemistry for white blood cells. $\endgroup$ Commented May 22, 2019 at 11:34
  • 7
    $\begingroup$ @Ishi: The author is a writer and editor, not a scientist: "Benjamin Radford is deputy editor of Skeptical Inquirer science magazine and author of six books including Scientific Paranormal Investigation: How to Solve Unexplained Mysteries." Absence actual scientific evidence, this article is not worth the paper it was written on. $\endgroup$ Commented May 22, 2019 at 11:53
  • 4
    $\begingroup$ @JMac: Haemochromatosis is specifically too much ferritin, a particular iron compound used by the body to dispose of iron. This disposition is done through elimination of the white blood cells that create ferritin. Thinking that haemochromatosis is too much iron kills people - like my dad, whose physician refused to listen to experts such as Paul Adams: Honorary Fellowship for Distinction - Royal College of Physicians of Ireland 2013. You are thinking of iron overload, the generic syndrome. $\endgroup$ Commented May 22, 2019 at 17:39
  • 7
    $\begingroup$ #2 and #3 don't seem particularly relevant. Yes, you can contract blood-borne diseases by drinking blood, but you can also contract water-borne diseases by drinking water. The fact that some blood is infected with disease doesn't make it an unsuitable replacement for water. And water contains no vitamins, and single-digit %DV of many minerals, depending on the source - I don't see how you could become malnourished by replacing part of your diet that has marginal nutritional value to begin with. $\endgroup$ Commented May 22, 2019 at 19:34

Here is a specific account of a person surviving at sea drinking turtle blood, while eating some fish and drinking some rain water.

Several peoples have a habit of drinking raw animal blood, at least for ritual purposes; here is a recent account. They probably drink more than a few spoons full so that the Lifescience article quoted by Ishi appears alarmist.

  • 1
    $\begingroup$ It is very unlikely that blood from any animal, including turtles, will be hydrating and we can't just rely on a story. $\endgroup$
    – Jan
    Commented May 22, 2019 at 14:46
  • 3
    $\begingroup$ @Jan Well, obviously it depends on how much water you ingest from other sources, and perhaps even on the food (because there is metabolic water). The body will also adapt to not having enough water (perhaps even by using stress metabolic pathways, like excreting uric acid?); the numbers found are surely the optimal, healthy ones. Apparently, blood is a net water drain under normal conditions, but whether that holds if there is little water overall is not immediately clear to me. $\endgroup$ Commented May 22, 2019 at 15:09
  • 1
    $\begingroup$ @Jan I also find refuting concrete evidence with "we can't just rely on a story" unsatisfying. What is science if not the collection and interpretation of stories? Of course ideally stories which have been intentionally and carefully produced, but stories nonetheless. $\endgroup$ Commented May 22, 2019 at 15:11
  • 8
    $\begingroup$ The first source you linked says that "We're not sure just how long you can survive off turtle blood, but Ivan did say that he was also drinking rainwater." We do not know the amount of rainwater he drank, so we cannot conclude that drinking turtle's blood helped. We can rely on stories and learn from them, when they clearly say what exactly happened, but we do not know this here. $\endgroup$
    – Jan
    Commented May 22, 2019 at 15:14
  • $\begingroup$ Very informative and the answer could then be yes indefinitely. $\endgroup$
    – Muze
    Commented May 23, 2019 at 22:30

You'd have no problem substituting blood for water, provided:

  1. You allowed it to coagulate first, and
  2. You prevented it from evaporating

Fresh blood has a high concentration of protein, certainly, but it's not a homogeneous fluid. If you decanted the blood into a container and let it sit, it would soon coagulate, causing the majority of the cellular and protein components to precipitate and stratify.

The plasma fraction (approximately 55% in healthy mammals, and not radically different in most species with hemoglobin) forms a supernatant (the top, easily accessible layer) which is 95% pure water by volume and about 5% protein.

The sodium salt concentration of plasma is about 140 mEq/L, or 3.2 g per L (mEq = mg / atomic weight * valence, Online Calculator). Since healthy kidneys can concentrate salt to 4-5X osmolality (and since the source material is physiologic blood), you'd have no problem drinking the top, liquid portion. You would also avoid most of the long term iron issues as the hemoglobin is found in the erythrocyte (deposited) fraction.

Thirsty vampires will be disappointed that plasma appears amber, not "blood red"

  • $\begingroup$ @Jan, total osmolality for blood plasma is 275-300 mOsm/kg - see sciencedirect.com/topics/medicine-and-dentistry/… for a reference. This encompasses salts, glucose and soluble proteins. $\endgroup$
    – Kvothe
    Commented Jun 6, 2019 at 19:37
  • $\begingroup$ @Jan, not clear what you mean by net water. If you take a 1 L blood volume, 55% of that (the plasma) is 550 mL. Is that what you mean? $\endgroup$
    – Kvothe
    Commented Jun 6, 2019 at 19:38
  • $\begingroup$ @Jan - I'm not sure the back and forth is helpful, but I think you're missing that the kidneys serve to actively concentrate fluid. Remember, your kidneys are actively concentrating your entire blood plasma volume about 60 times per day (which just like the orally consumed plasma is about 300 mOsm/L). Kidney input is 300 mOsm/L, and output can go to 1500 mOsm/L. For example, the original WHO Oral Rehydration Solution was 311 mOsm/L WHO ORS and rehydrates. Also the "protein" in plasma is predominantly urea already. $\endgroup$
    – Kvothe
    Commented Jun 7, 2019 at 21:14
  • $\begingroup$ @Kvohte, urea needs to be removed from the blood by urine. Consumption of 25 g protein (5% of protein in 500 mL of plasma) will result in 25 x 8 = 200 mL of urine, because 1 gram of protein results in the amount of urea that needs 8 mL of water to be excreted ( nap.edu/read/10925/chapter/6#135 ). This means that by drinking 500 mL of water from plasma, you get only 300 mL of net water because the urea will force 200 mL of it out into the urine. Do you agree with this? So, plasma would be still hydrating, but not as much as it seems. I need to think about sodium effect a bit more. $\endgroup$
    – Jan
    Commented Jun 8, 2019 at 10:30
  • $\begingroup$ Can't believe this question was accepted, blood literally kills you if you drink it... $\endgroup$
    – user59946
    Commented May 19, 2020 at 23:01

You must log in to answer this question.

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