The American Academy of Pediatrics does not recommend water safety lessons for babies less than one year old:

The water-survival skills programs for infants may make compelling videos for the Internet, but no scientific study has yet demonstrated these classes are effective

However, many parents subject their babies to "survival swimming lessons", believing that these will decrease the child's risk of drowning. In the United States, an organization names Infant Swimming Resource (ISR) is a major provider of these lessons. Their classes for babies 6-12 months old focus on teaching them to roll over and float on their backs:

Generally speaking, children ages 6 months to 1 year learn the ISR Self-Rescue® skill of rolling onto their backs to float, rest and breathe. They learn to maintain this position until help arrives.

In their marketing materials, ISR providers claim that after 6 months old, "the sooner the safer", that infant swimming lessons save lives, etc.

Is there any evidence to suggest that this is true - that teaching infants who are 6-12 months old to back float reduces their risk of fatal drowning?

(Note that I am not asking whether babies are natural swimmers; I am asking whether teaching them to roll over and float on their backs is proven to be effective in reducing their risk of drowning.)


2 Answers 2


No, there is no scientific evidence and that is explicitly stated by the AAP link you mentioned. The studies referenced by the press release (2010) are preliminary at best and most are methodologically invalid. I'm unaware of any properly designed study.

  • 1
    This answer basically says "Take the AAP's word for it (and my word that the AAP is correct, and that nothing new has come up in the last six years)." To a skeptic, that's not a very useful argument. A better answer would show us the evidence and enumerate its flaws.
    – ff524
    Commented Oct 22, 2016 at 23:45
  • @ff524 If there are no evidence, are you expecting me to prove a negative or what? The critique of the studies already conducted is discussed in the very link in the beginning of the question. I'm still unaware of others. Please post a different answer and downvote this one if you can come up with something contrary to the current AAP's position. I couldn't and there's no point in being skeptic just for the sake of it. One withholds believe until there's evidence, not the other way around. Commented Oct 22, 2016 at 23:54
  • 1
    @CelticWarrior: Skeptics.SE is unusual by its very nature, in that the burden of proof falls on the answerer, where it really belongs on the original claimant. Proving a lack of evidence is difficult, and the proxy we tend to use (even though it is an Appeal to Authority) is to cite someone who has the expertise, and has searched the literature and said they came up empty-handed. In this case, you have effectively done that, but the concerns seem to be whether they are suitable experts.
    – Oddthinking
    Commented Oct 23, 2016 at 21:04
  • I trust them (AAP) tentatively. I trust their European counterparts more, TBH, and they aren't at odds in this issue as well as the vast majority of issues. There are a couple of exceptions where I find the AAP more concerned with their members' income than the children best interests. Commented Oct 23, 2016 at 21:16
  • "I'm unaware of any properly designed study", are you claiming that based on evidence or just as a personal opinion? Did you look for better studies?
    – Sklivvz
    Commented Oct 24, 2016 at 8:40

This answer is incomplete, but hopefully contains some relevant and helpful information.

Globally, the highest drowning rates are among children 1-4 years

  • Australia: drowning is the leading cause of unintentional injury death in children aged 1-3 years.
  • Bangladesh: drowning accounts for 43% of all deaths in children aged 1-4 years.
  • China: drowning is the leading cause of injury death in children aged 1-14 years.
  • United States: drowning is the second leading cause of unintentional injury death in children aged 1-14 years.

There are some issues here. We don't know why the groups start at 1 year. Given that infants are immobile for the first 5-7 months of their lives, there are too many possible reasons why the statistic starts at 1 year. So we look for other sources: enter image description here

We see that percentage wise, the death by drowning rate for infants less than a year old is low. But that seems to be because too many of them suffocate, so let's try to somewhat normalize the data. Number of deaths "n" was 5883, 7% is 412. For the 1-4 year olds, number of deaths was 10203, 27% is 2755. Let's assume the 1,2,3,and 4 year olds are equally likely to drown, which gives us 689 deaths for each year. This seems to be comparable to the 412 for less than 1 year olds, especially considering that less than 1 year olds can't move for the first 5-7 months of their lives.


  1. For 1-4 year olds, risk of drowning is significant, compared to other causes of unintentional injury death.
  2. For 6-12 month old infants, the risk of drowning seems to be comparable to that of 1-4 year olds.
  3. For ages 5 and above, risk of drowning is significantly lower than for ages less than 5.

So, where do children drown? [pdf]

enter image description here


Children under one year most often drown in bathtubs, buckets, or toilets. Among children ages 1 to 4 years, most drownings occur in residential swimming pools. (Where children drown, 1995, Pediatrics 2001;108(1):85–9)

Due to lack of space, backfloating will not save someone from drowning in a bucket or a toilet. Knowing how to backfloat certainly does increase survival chances once inside a pond, bathtub or swimming pool, since not dying in these situations is really all that backfloating is.

This answer is incomplete because I do not have any data about the link between the reduced chance of drowning in pools and bathtubs, and the possibly increased exposure to any kind of drowning risk after being taught how to backfloat, and reduced risk awareness of the parents when they know their infant can backfloat.

  • 1
    This does not answer the question
    – user22865
    Commented Oct 24, 2016 at 13:06

You must log in to answer this question.

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