Many people ask if it is possible to inhale acrolein from decomposed glycerine in e-cigarette vapour. An opinion on this is below (and note that it is an opinion). Short version There is no indication that current production ecigs produce any more than trace amounts of acrolein since this has been tested for under correct working conditions, with no positive results. Non-standard equipment, though (e.g. sub-ohm RBA use), has no evidence base for any opinion - there are no tests. It is very easy to heat e-liquid in a lab test to produce acrolein, in conditions that would be impossible to replicate in a real e-cigarette in use by a real person not a machine - (a) a person would choke on the smoke; since (b) at this point the aerosol product has become smoke, not vapor. Any 'study' that purports to have located acrolein in measurable quantities in regular ecig vapor (as against sub-ohm RBA vapor, which is a different issue) did not use a realistic test set-up - we already know that acrolein cannot be detected in ecig vapor from a regular atomiser tested correctly. We already know the true facts (not those cooked up in research funded by commercial rivals). The biodiesel byproduct glycerine issue is therefore probably more important; this is explained below. Nothing at all is known about the vapor products from sub-ohm rigs and it may be a mistake to assume they are the same as from regular atomisers / cartomisers / clearomisers. Explanation When glycerine is overheated  and it decomposes (i.e. it is 'burnt'), acrolein is produced, which is a toxic chemical. This means that there is a genuine reason to examine the possibility that acrolein may be present in inhaled vapour from an e-cigarette, since almost all refills contain glycerine (aka glycerol, glycerin, VG). The percentage of glycerine may be as much as around 75% for '100% VG' liquids (as they can be about 75% glycerine plus 10% DW distilled water plus 15% flavouring) to ~20% or less (a regular e-liquid mix could be 70/30 PG-VG plus flavouring, therefore about 63-22-15 PG-VG-flavour). The highest amount of glycerine in an e-liquid refill could be about 90% (although this would be unusual as the material would be highly viscous - thick and gloopy), the lowest zero to 5%. The term 'VG' is often used because in the past, vegetable-source glycerine was the best choice for vaping as the choice was between animal carcass source, biodiesel byproduct source, or vegetable source (palm oil or coconut oil). There was no easily-sourced synthetic glycerine. Why we don't inhale acrolein There are multiple reasons we don't inhale acrolein in vapour in anything except trace amounts : 1. An atomiser is a liquid-cooled heater element. In order for the atomiser coil to be hot enough to produce acrolein, there must be no liquid to cool it. If there is no coolant, then the temperature can rise and the coil can become very hot. It may then become red hot; but if there is no liquid, then no acrolein can be produced. 2. Adding anything to glycerine reduces its boiling point. This means that the glycerine+additive mix will boil off well below 280 C, and in fact the boiling point of a mix may be as little as half this. (Water has a much lower boiling point and PG is only 188 C.) 3. The negative pressure and air throughput within an atomiser may also increase the cooling effect and reduce the boiling point. 4. Acrolein will probably not be produced in any significant quantity in a regular atomiser until the smoke phase is reached, which is well past the point at which the vapour becomes impossible to inhale. We know this because of the multiple tests run by reliable investigators on regular ecig heads. These heads are all of micro-coil type enclosed in small-diameter tubes, with silica wicks. We have no data for other types of head. 5. RBA use is a different matter as no tests have been carried out on this type of atomiser, which is known to run hotter under certain circumstances, and may have different vapor production characteristics. See also: http://www.e-cigarette-forum.com/forum/health-safety-vaping/638690-formaldehyde.html#post14957231 Acrolein smells nasty It is the acrid smell that you get from the smoke when you burn cooking oil in a pan (as that is acrolein). A very small amount would be easily detectable in ecig vapour. This is why you do not continue as soon as a dry vape is experienced, as the next draw will be burnt, and a burnt taste equals smoke, and smoke equals toxic materials. There is an interesting feature of laboratory tests of e-cigarettes where the protocols used were faulty in some way, since smoke is an inevitable result of faulty vaping protocols. A human user would not be able to inhale such materials, but a test rig cannot tell the difference between perfectly moist vapour and hot, dry, toxic smoke. For example, many lab tests that provided photos of the test rig show the e-cigarette operated inverted (upside down) or at an angle with the tip high, in which position it is almost guaranteed to dry out and burn up (as an e-cigarette is a gravity-fed liquid-cooled heating device, much like your electric kettle - and they don't work upside down). Lab tests of this kind would be very likely to produce anomalous results, as the protocols are critical (angle of operation, puff length, inter-puff interval duration, max # of puffs, puff volume, etc etc) and are radically different from tobacco cigarette testing protocols. Detection of acrolein would be a typical example of a faulty lab test such as this. There is probably a considerable variation in quality between the multiple lab tests of ecig vapour. For example, Intellicig UK were the first large-scale users of glycerine as the main e-liquid excipient, and have extensive laboratory test facilities including oversight by qualified academics from their local university and further afield. They ran tests at up to 300 C and detected no measurable quantity of acrolein from their micro-atomisers, as used in the M401 and similar. Such tests are probably repeated occasionally since they are one of the more conscientious vendors, and also have more to lose than most vendors due to their scale. It is probably reasonably safe to say that if they do not consider there is an issue, then regular ecigs probably have little risk . Is acrolein the biggest issue anyway? The acrolein question is probably not the main glycerine issue at this time, in any case - because it is very unlikely that an ecig user can inhale anything other than the most minute quantity of acrolein, and almost unmeasurably small quantities of anything are not toxic. The main issue currently is the biodiesel byproduct problem; specifically the jatropha plant issue. Biodiesel production involves the use of multiple plant sources and these include jatropha in the modern era. Glycerine is produced as a by-product, along with diesel fuel. The problem is that jatropha is toxic to humans: its esters are carcinogenic. For this reason biodiesel by-product glycerine absolutely MUST NOT be used for vaping-use glycerine. However this is easier said than done: no one really knows if this condition has been met unless the finished retail product is tested for the carcinogenic phorbol esters of the jatropha plant. As hardly any e-liquid vendors either test properly , or know how to test for this contaminant in any case, it is a genuine and serious issue. The FDA have warned about this and a web search reveals multiple resources, including: Potential Toxicity Associated with Oils, Glycerin, and Proteins Derived from the Jatropha Plant How to obtain jatropha-free glycerine Because the supply chain in the e-liquid world is based on word of mouth ("My supplier told me this is the highest quality") and overseas certificates of dubious quality, it is impossible to eliminate the possibility of contaminated glycerine unless (a) a reliable test certificate is presented, or (b) the source is unimpeachable. A 'reliable test certificate' is not one that originates overseas, it comes from a nationally-accredited testing laboratory in your country, and it states specifically on it that jatropha was one of the contaminants tested for. Alternatively a reliable source is needed, such as a pharmaceutical supplier who can provide (not tell you about) pharmaceutical licenses for inhalation duty; the safest at this time is probably Dow Chemical's Optim product, a synthetic glycerine of about 99.9% purity (nothing is 'pure'; the contaminant here is likely to be water). This has multiple licenses for inhalation and Dow advise its use for medical inhalation in preference to their pharma grade PG (which used to be the most popular for e.g. asthma inhalers, but is now losing out to glycerine as there is no drying-out of the throat and upper lung area that some experience with PG). E-liquid manufacturers can obtain Dow Optim in 4-gallon containers in the USA. It is not cheap - but there is no good, cheap e-liquid since this is impossible (testing is expensive, good materials are expensive, and qualified chemists are expensive to employ). A practical alternative, to be used until such time as manufacturers provide full test details or are regulated by government in order to prevent the inclusion of contaminants (very likely and probably desirable if the industry will not solve the problem themselves), is to obtain your glycerine-based e-liquid from a manufacturer with a multi-million turnover and a very good reputation. This is because (a) regular full testing is expensive and small-scale vendors simply cannot afford it unless they are members of a trade association that handles the testing for them at a discount, and (b) a large-scale vendor with a very good reputation that would be completely destroyed if they were exposed for selling contaminated materials is a good bet, since they will probably be taking steps to ensure such an incident does not occur as they have a lot to lose. Conclusion Inhalation of acrolein is unlikely to be a problem with regular ecigs (minis and mid-size models). It is impossible to remove trace contaminants, but these are not seen as clinically significant here since they are universally present in all consumer products and impossible to remove completely. Lab tests that show measurable quantities of acrolein are very likely indeed to have been conducted using protocols that do not represent normal usage, or in RBAs run hot. No study declaring that acrolein was produced by a regular e-cigarette has ever been supported by independent investigators. Some would say, though, that testing is not required to validate any opinion: most people can reliably detect the difference between wonderfully flavoured water vapour and toxic smoke. If you are not sure of the difference, burn up some cooking oil in a frying pan and get it really smoking, place your head right over it, then take a deep breath. That's acrolein. [p.s.: don't do this] Maximum risk reduction is something that many people will consider pointless, although it may be of interest to some. For example, it seems unlikely that any contamination of e-liquid within normal parameters could be expected to elevate risk beyond about 99% less than that of smoking; it seems a reasonable bet that no matter what level of contaminants is present, within a reasonable range, vaping cannot be anything other than so much safer than smoking that there is no comparison. But note: - It is possible that vaping could be made 10,000 times safer than smoking, assuming you felt it worth the trouble of examining every factor for risk and then taking steps to reduce any perceived elevation of risk - It is even possible that vaping might perhaps be made as high-risk as almost 1/20th or even 1/10th the risk of smoking, if you wish to indulge in extreme vaping while using the cheapest hardware and refills Nobody knows the answers to such questions, and several decades will need to elapse before things are clearer. --------------------------  The temperatures at which glycerine boils or decomposes producing acrolein are in a similar range, around 280 - 290 Celsius.  It is impossible to remove trace amounts of contaminants from anything/everything that you breathe, drink, eat, or touch. The dose makes the poison, so trace amounts are not significant especially in modern urban life, where overall risk for such things is far higher than rural life in an agrarian economy. There is a price to pay for the modern urban lifestyle, and constant exposure to minute amounts of contaminants is one of the costs; there may or may not be a health impact from such continual exposures.  Of course, we are talking about a mini ecig here, with low performance by modern standards. We are not talking about a sub-ohm RBA generating massive clouds of hot vapor. If you use such a device then you must expect to be way out beyond the bleeding edge; there may be an increase of risk in one or more departments. There might be a considerable elevation of risk, or there may be none - we have no idea at all, currently.  If you wish to contest this, then please direct me to the websites of e-liquid producers who have published recent full tests of finished retail products and provide the full details of such tests; or who publish a current certificate from an authorised laboratory who state that the feedstock glycerine was tested for jatropha esters. If this information is widely available I am very happy to change the statement; or even to provide exceptions for those who currently comply. At 2013-08-01 I found none who complied, who I would be prepared to recommend as safe. This is an industry issue that needs resolving or someone may resolve it for us.