There are a number of ways to purify water, off the top of my head we have filters, iodine, and boiling. Clearly each has some advantage/disadvantage (iodine tastes terrible, boiling takes time). Which of these is the safest? Is there another method that's safer?

  • 1
    related: outdoors.stackexchange.com/questions/1117/…
    – user2169
    Commented May 3, 2013 at 16:43
  • 5
    Does it really need to be the safest way? Make it "safe enough," striking a balance between the risk of getting sick and purifying it to ridiculousness. Don't get carried away by thinking it has to be perfectly pure. It never will be, and that's okay, except for people with unusual medical conditions. Commented Aug 6, 2013 at 22:40
  • 2
    Save water, drink beer.
    – Andrew
    Commented Nov 28, 2014 at 15:53
  • 2
    It might be helpful to clarify that you're talking about purify water from diseases, not from poisons/chemical contaminants.
    – Erik
    Commented Mar 9, 2016 at 21:47
  • The safest way would be to use an autoclave (en.wikipedia.org/wiki/Autoclave). As you can see this is not very practical, but you did ask for the safest way...
    – fgysin
    Commented Oct 3, 2016 at 6:00

10 Answers 10


When you're asking for the safest way to purify water, you're asking for the method that removes the most harmful stuff from the water, like bacteria, viruses and larger impurities like mud or sand.

No one method is really perfect at removing everything, so I usually use a two-stage approach:

  • Filter: If the water source is cloudy, your first step should be to remove the larger impurities. You can buy expensive pump filters for this, or you can just pour the water through a sock, bandanna, or any other tightly-woven fabric to remove the larger stuff.

    Note that you can probably skip this step if the water you're collecting is clear and fast-moving.

  • Boil: Heat sterilization is really the only sure-fire way to kill bacteria, viruses and other undesirable microbial life in your water. It's a commonly used technique to clean medical instruments since it's so effective.

    It does require a heat source, and it'll take some time to get a full rolling boil. But if you're at all worried about the quality of your water, it's your safest bet.

Edit: As commenters noted, if you're in an area with potential chemical pollutants or metals in the water, you should filter even if the water appears clear. But if you're really in a situation where mother nature's water has been tainted by man-made waste, you should probably pack in all of your own water anyways.

  • 6
    I wouldn't rely on whether the water is clear and flowing. A carcass up river could be rotting away without discolouring the water. I'd always boil water I collect (if possible). Commented Jan 24, 2012 at 20:49
  • 5
    Also - seemngly clear water can still have a high metal content - drinking arsenic and lead laden water is not particularly advisable.
    – HorusKol
    Commented Jan 24, 2012 at 22:46
  • 4
    "It's a commonly used technique to clean medical instruments" Just wanted to mention that this method also requires pressure and steam. Don't expect medical sterilization just by boiling. Water can't get past 100 Celsius at standard pressure. en.wikipedia.org/wiki/Autoclave
    – rtf
    Commented Aug 27, 2012 at 20:31
  • 4
    It's not true that heat sterilization is the "only sure-fire way" to kill all disease-causing microorganisms. UV treatment (steripen) also kills all types of microorganisms (bacteria, viruses, and protozoan cysts). Unlike boiling, UV treatment is practical for water collected during the day while hiking.
    – user2169
    Commented Apr 30, 2013 at 22:32
  • 5
    "where mother nature's water has been tainted by man-made waste" - arsenic and other heavy metals can (and do) occur naturally, so stating "man-made" here promotes the false idea that "natural" is safe
    – HorusKol
    Commented Jun 6, 2013 at 14:09

I would get a ceramic filter that is rated to remove metals like arsenic (this is a particular issue in the UK where most surface water in the wild country can harbour arsenic - but I'd rather assume all ground water to be 'tainted' than drink it and become poisoned). I've used some hand-pumped filters which can draw water from a lake or river, and filter at the rate of about a litre a minute.

Then, drop an chlorine tab into the water - this usually requires 30 minutes to completely kill off biologic contamination. You can then get neutralisers to remove the taste of the chlorine from the water.

Boiling shouldn't be necessary after this treatment - unless you want a cuppa.

One thing - do not rely on the apparent clarity of the water - bacteria/viruses and dissolved metals will almost always be invisible to the naked eye.


There is typically no need to purify water collected from natural sources in the wilderness. For example, in a survey of 69 sites in the U.S Sierra Nevada, every site had concentrations of Giardia cysts much too low to make anyone sick.[Rockwell 2002] The perception that backcountry water is unsafe to drink without treatment is folk wisdom that is controverted by the available scientific evidence, at least in the backcountry areas in the US such as the Sierra that have been extensively studied. Of course none of this applies outside of the backcountry, e.g., you certainly don't want to drink water in Arkansas that comes from runoff from a poultry farm.

When people do actually contract backpacker's diarrhea from exposure during a hiking trip, by far the most common reason is hand-to-mouth contamination.[Welch 1995] The most effective disease prevention measures are to wash your hands after pooping, refrain from sharing pots and pans, and use freezer-bag cooking so that food never goes in your pots.

If you do want to purify your water, there's a lot to be said for UV (steripen), which is one of the few methods that will kill all three types of disease-causing organisms: viruses, bacteria, and protozoan cysts. Unlike boiling and halogens, it gives you water that you can drink immediately while hiking during the day.

For more detail, I have a long article on this topic here, with lots of citations to reliable scientific and medical sources: Drinking Water for Hiking: Myths and Facts. I realize that my answer goes against the conventional wisdom, and that's why I've provided references to scientific papers in this answer, as well as much more extensive references in the longer article.

Rockwell 2002 - Robert L. Rockwell, Sierra Nature Notes, Volume 2, January 2002,

Welch 1995 - Thomas R. Welch and Timothy P. Welch, "Giardiasis as a threat to backpackers in the United States: a survey of state health departments," Wilderness and Environmental Medicine, 6 (1995) 162.

  • 12
    Perhaps what you say is true for the Sierras, but is certainly NOT true of other areas. Various places in the White Mountain National Forest of NH were tested, and enough giardia found that there is a overall recommendation not to drink the water directly. If giardia is the only issue, then even filters work since these buggers are rather large. I am surprised you were the only one to mention germicidal UV. It seems like a really good idea to me. 90 seconds for a liter and you're ready to drink. By the way, I wrote the firmware for the latest SteriPen, the Ultra. Commented May 1, 2013 at 21:50
  • 2
    I looked around a bit but couldn't find a definative source. Back in the early 1980s everyone pretty much used backcountry water in the WMNF without thinking much about it. By the late '80s that was not done anymore. Giardia had been found in several fairly deep backcountry sources, so one had to assume it could be in any water, even if it actually wasn't in most. The one I remember specifically was a spring on Signal Ridge, on the south shoulder of Mt Carrigain, but there were others. Since giardia can pop up quickly, it's smart to assume any water could be infested. Commented May 4, 2013 at 22:00
  • 3
    @BenCrowell: as per my answer - it's just safer to assume tainted sources and work from there, you're not going to degrade the water quality. A hand-pumped ceramic filter is pretty light-weight, and purification tablets are negligible.
    – HorusKol
    Commented Aug 21, 2013 at 5:01
  • 2
    @HorusKol: I see. I hadn't realized that arsenic could be removed by ceramic filters. The big issue IMO is simply that people are so hung up on treating their water, which is typically pointless, that they completely fail to pay attention to the real issue, which is hand-to-mouth contamination.
    – user2169
    Commented Aug 21, 2013 at 5:05
  • 2
    Neither Ben Crowell, Robert Rockwell nor Thomas Welch are epidemiologists and their conclusions are refuted by the CDC. appalachiantrials.com/… For pathogens there is no better method than boiling. Most consider this impractical, however. To be really careful, a combination of a good filter combined with chlorine dioxide would be really hard to beat and provides redundancy. For backcountry in North America, most knowledgeable people feel comfortable with either a good filter or chlorine dioxide alone. Commented Mar 13, 2016 at 19:22

If safety is your primary concern, then the only two methods can guarantee safety from major contaminants (microbial or otherwise) these are Distillation and Reverse Osmosis


Boiling alone will leave most contaminants in the water, such as heavy metals, hydrocarbons, etc... multi-stage distillation will both kill pathogens and leave most contaminants in the residue.

Distillation can be carried out using a solar still if a fire and appropriate equipment isn't available. One stage distillation is usually enough unless the water source is heavily contaminated.

Reverse Osmosis

Is usually combined with another method of filtration/purification, but has very high water purity if used properly, although the energy required and the efficiency when using low pressures (like a hand pump) make this a poor choice when hiking.

Ensuring safety

In both cases, the water really should be filtered until clear beforehand for the best results and ease of use, although this is not strictly necessary.

For the best results, purifications/filtration methods should be used together, for example, activated carbon filtration of distilled water (to remove volatile organic copounds) or chlorination of water recovered by reverse osmosis.

Distillation and reverse osmosis are also the only two methods mentioned so far able to desalinate seawater and recycle urine.

For example, The ISS uses a low pressure distilation system to recover drinking water.

You will need to re-mineralise the water in both cases, both for taste and electrolytes.

Both methods require a lot of energy, but will guarantee exceptionally safe water.

  • 1
    I don't see how this applies to the topic of this site. A still may be practical about the space station, but ...??
    – user2169
    Commented Feb 10, 2017 at 18:03
  • @BenCrowell, distillation and reverse osmosis are suitable for the outdoors. For example, expeditions with > ~12ppl make the use of a reverse osmosis hand pump worth the weight and energy requirements if uncontaminated drinking water is an issue. Solar stills are also a great addition to desert or marine expeditions, and understanding the principles of a solar still means that you can extract water out of places that would otherwise be inaccessible, (toxic plants, spoilt food) with nothing but sunlight and a plastic tarp.
    – brice
    Commented Feb 12, 2017 at 1:53

Boiling is the best thing to do, as stated already - I was always taught to boil for 2 minutes and then it's safe. (Far from an accepted time though. For debate on how long you should boil, see here: How long does water need to be boiled for to kill all bacteria / viruses?)

As pointed out below though it may be wise to use a filter to get rid of any toxic metals prior to boiling. I've personally not had an issue with them but it depends heavily on the area, and better safe than sorry!

In situations where you can't boil it then chlorine tablets can be an alternative; while I don't have a source to hand I'm sure that I read they're generally recommended above iodine tablets because the latter if used extensively over time can damage the thyroid.

  • 2
    Boiling wouldn't remove metals
    – HorusKol
    Commented Jan 24, 2012 at 22:55
  • @HorusKol Good point - an oversight on my behalf. I've updated the answer accordingly.
    – berry120
    Commented Jan 24, 2012 at 23:01
  • 1
    @BenCrowell A hotly debated point - I've seen people argue for it both ways. Personally I prefer to boil for a bit longer because I've never had any trouble doing that, and don't really see the need to cut down with most of the hiking I do. I agree it may be safe to boil for less time though.
    – berry120
    Commented Apr 30, 2013 at 22:43
  • 1
    @BenCrowell Sure, but almost everyone who has an opinion on the matter will throw some form of "evidence" to back themselves up. I'd call your source into question on a number of points, for instance is raw milk often swarming with the same type of microbes that can be in backcountry water, or is it a more restricted subset that can't survive over 72C? Official guidelines also disagree with your assessment (water.epa.gov/drink/emerprep/emergencydisinfection.cfm#method), they recommend boiling for one minute. Of course, some microbes can even survive boiling indefinitely...
    – berry120
    Commented Apr 30, 2013 at 23:22
  • 1
    @berry120: "Official guidelines" are not scientific evidence; they reflect a bureaucratic/political/social consensus. Your statement that "some microbes can even survive boiling indefinitely" is literally correct, since there are extremophile bacteria that can live at temperatures above 100 C. However, there are no such organisms that are known to cause disease in humans when drunk in water. If you claim there is some organism capable of causing disease in humans that can survive 15 seconds at 72 C, please tell us what it is and give some scientific evidence that it has that ability.
    – user2169
    Commented May 2, 2013 at 16:00

If boiling isn't an option, aqua mira is usually a good choice. It's a doddle to use, and is very effective. It doesn't make the water taste like soap either, like some of the tablets do.

  • 1
    Halogens such as aqua mira aren't very effective against protozoan cysts such as Giardia and Cryptosporidium. The real question is what disease-causing organisms you want to protect against. In backcountry areas in the US that have been well studied, the answer is that there simply aren't any organisms, of any type whatsoever, in concentrations capable of causing disease. (For scientific references, see my answer and the longer article linked to from it.)
    – user2169
    Commented Apr 30, 2013 at 22:29
  • 1
    @BenCrowell what about those backcountry areas that haven't been well studied? Do you have a map of the well studied areas?
    – HorusKol
    Commented Jun 6, 2013 at 14:14
  • @HorusKol: My own answer gives a link to a longer article I wrote. The article gives all the scientific references I was able to find. A lot of the water-analysis results refer to the Sierra, and most of the epidemiology was done for the US. It would be interesting, for example, to see specific studies for the Alps, which have more permanent human settlement and more cattle farming.
    – user2169
    Commented Aug 20, 2013 at 17:48
  • true. i'm in NZ and i suspect a lot of the back country water away from animal farming is drinkable straight out the river, but i still use mira as a backup
    – Sirex
    Commented Aug 20, 2013 at 19:45
  • @BenCrowel said "In backcountry areas in the US that have been well studied, the answer is that there simply aren't any organisms, of any type whatsoever, in concentrations capable of causing disease." Absolute non-sense: "Although the advice to universally filter and disinfect backcountry drinking water to prevent disease has been debated, the health consequences of ignoring that standard water treatment advice have been documented" CDC bucktrack.com/water.html Commented Mar 14, 2016 at 15:57

I have no science to back me up, but the SAFEST way would be boiling. Since water all boils at the same point; you know as soon as it starts rolling that the nasties are all being killed. Hold at a rolling boil for 2 minutes (as I was taught), and it is safe to drink.

  • 1
    I recently read it should boil for at least 10 minutes to be sure all the "nasties" are dead.
    – Noam Gal
    Commented Jan 24, 2012 at 20:25
  • 1
    Wow. I wasn't sure if it should be 1 minute or 2 and had to run, so I picked the longer. If something can survive 2 minutes at a boil, I guess I'm ok with it killing me, it's earned the right.
    – Pulsehead
    Commented Jan 24, 2012 at 21:08
  • mm - You were probably right, I just looked for a reference for that 10 minutes claim, and found this - jolly-green-giant.blogspot.com/2009/03/… - it says 1 minute is enough in most places, up to 3 over 6500 feet altitude. Sorry for the wrong info.
    – Noam Gal
    Commented Jan 24, 2012 at 21:13
  • 2
    Boiling wouldn't remove metals
    – HorusKol
    Commented Jan 24, 2012 at 22:55
  • 1
    Bringing it to a boil is sufficient at all altitudes. It is not necessary to boil for any amount of time. See my comment on berry120's answer and the longer article linked to from my own answer.
    – user2169
    Commented Apr 30, 2013 at 22:22

Fecal contamination should not really be an issue unless someone is already sick. Recent research has suggested that even eating ones own poop should not make you sick. Fecal transplants is a new way of treating some issues that form from a lack of flora in your gut. Can You Eat Your Own Poop? I for one would use something to treat the water. My cousin hiked the Sierra MTS a few years ago and contracted giardia so just from that knowledge I would definetely treat the water with something.


This is an excellent article from from REI.
Your choices include:

  • Pump-style filters and purifiers
  • Gravity-fed filters and purifiers
  • Ultraviolet light pen purifiers
  • Sip/squeeze bottles with in-line filtering straws
  • Chemical tablets
  • Boiling

enter image description here

Boiling alone is very effective the problems that is takes a lot of time. You pretty much need to do it at camp.

Another factor is filtered water tastes better. Chemicals make the water taste worse.

Have a filter (or one of the other options) for when you need water on the trail. Leaving camp with 3 liters of water is 6.5 lbs. Viruses are rarely found in North American wilderness waters.

As for heavy metals. If there are heavy metals in the water then not reason to go hiking there. Heavy metals are not naturally occurring in the wilderness.

Bring a collapsible water bucket for camp so you don't have to multiple trips. water bucket

Not yet mentioned is lots of large particles in the water.
enter image description here

You will fowl the filter faster with large the particles. Most filters backwash to restore most of the flow but still better to get rid of the big stuff first.

If you boil you still don't want to drink water that looks like that. And it does not taste good.

Just put the water in a vessel and let the large particles drop to the bottom. A good share will drop in 2 hours and after 8 you are not going to get much more drop out.

For cooking water you can use it direct as you are going to boil it anyway.

Bring a small siphon tube and place the container high enough the water bottle is below it.


This Discovery News article says:

A simple and inexpensive — but not necessarily the best tasting — method of purifying wild water is by dropping in a couple of purification tablets or drops. The most common chemical used is iodine, but chlorine or potassium permanganate are also effective. Let the chemicals treat the water for at least 20 minutes before consuming, and mix it with powdered mixes to mask any of its taste.

  • I edited your answer to add a link to the article from which you copied the text. I notice that you've now edited the answer again to remove that link. I'm a little confused as to why you did this: personally I find that answers which cite their sources are (other things being equal) better than ones that don't, so I thought that my addition improved your answer. Please could you explain why you removed the attribution?
    – Pont
    Commented Mar 10, 2016 at 21:07
  • 1
    RV - plagiarism is not permitted here. You must include attribution - as Pont has kindly done for you here.
    – Rory Alsop
    Commented Mar 11, 2016 at 14:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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