We know that HIV can't be transmitted by mosquitos, and nor can other highly virulent viruses that are transmitted through blood and bodily fluid exchanges, such as Ebola (thankfully!).

Marcus Junius Brutus asked an interesting question in a comment on that above-linked question's answer:

[Why can't] tiny droplets of blood remain on the proboscis of the mosquito long enough for said mosquito to inject its proboscis into the bloodstream of another person? In particular, in what physical aspects is the proboscis different than a needle or razor? I am not talking about blood sucked into the gut of the mosquito, I am talking about blood remaining on the proboscis

There's no doubt that dirty mosquito proboscises aren't a vector for blood-based viruses (if they were, Ebola would be completely uncontainable in West Africa), and I understand the explanation on the above linked answer that there's no route from the mosquito's stomach to its proboscis other than via saliva production, but given how tiny the amounts of blood are that are required for transmission of such highly virulent viruses, why does it never happen in a similar way to transmission from a dirty needle or razor?

I've never seen a mosquito put its proboscis in sterilising fluid or an autoclave between feedings...

My best guess would be that mosquitoes are very efficient at taking up all available fluid, and that surface tension means that at this scale it is possible to take up every last droplet - but that's just a guess.

  • $\begingroup$ I guess a spin on this question is: is one virus 'particle' enough to cause an infection (or what probability)? I understand that viruses are extremely small. $\endgroup$
    – Doddy
    Commented Apr 5, 2016 at 10:31
  • 1
    $\begingroup$ Doddy - I believe you're taking about virulence, which varies from pathogen to pathogen. Ebola and HIV I believe are extremely virulent, readily transferred via dirty needles etc, which is why I chose them as examples (since even they clearly aren't transmitted from a used probiscus) $\endgroup$ Commented Apr 5, 2016 at 11:30
  • 9
    $\begingroup$ Droplets might be the key to this. Apparently, the proboscis inner diameter can be something like 25 microns or so, while blood droplets are apparently around a milimeter or so (that's where the shape of the droplet is almost perfectly spherical, which means great surface tension). So it might very well be that no blood is "stuck" on the proboscis after a feeding. I'm quite interested in the actual answer, though :) $\endgroup$
    – Luaan
    Commented Apr 5, 2016 at 12:08
  • $\begingroup$ Always relevant videos for proboscis: one, two. $\endgroup$
    – marcelm
    Commented Apr 5, 2016 at 17:32
  • $\begingroup$ @Luaan I'm not an expert at all in this field, but I find it hard to believe that there's a 0% probability that just because a droplet of blood happened to be on the proboscis of a mosquito at one time means that it definitely cannot inject the virus into a subsequent victim. What does surface tension have to do with it? Can't virus particles stick to surfaces? $\endgroup$
    – Doddy
    Commented Apr 5, 2016 at 22:24

2 Answers 2


A mosquito's proboscis isn't like that of a butterfly, which could easily have nectar clinging to it when it is coiled up; instead, consider that the part of the mosquito's proboscis that enters a blood vessle is probably wiped clean when it is retracts outward through the epithelium. A a dirty needle or razor is many, many times larger than the sucking parts of a mosquito, and a razor is an unlikely vector of HIV.

Epithelial tissue consists of cells very tightly bound together; it has to be to prevent constant invasion by bacteria, fungi, etc. It takes some considerable force or sharpness to penetrate. If you've ever watched someone give you an injection (a vaccine, for example), the needle is never bloody upon withdrawal. You might get a bead of blood from the wound, but the needle itself will be quite clean. The reverse is also true due to essentially the same reasons: the likelihood of introducing a pathogen by dragging it from the surface into the subcutaneous area is so low that there is no absolute need to swab visibly clean skin before giving a subcutaneous injection:

Although skin that is visibly soiled or dirty must be washed, swabbing the clean skin of a patient before giving an injection is unnecessary. Studies suggest that there is no increased risk of infection when injections were given in the absence of skin preparation. (WHO Bulletin)

HIV is the most serious virus (aside from Ebola, about which much is still unknown) which can be transmitted by needlestick. Sharps injuries - needles, suture needles, scalpels, lancets, etc. - are very common in the medical field, estimated to be much higher than 350,000/year due to underreporting and occurrence in non-hospital settings, e.g. nursing homes, etc. Of course, in the majority of sharps injuries, the sharp wasn't likely to have been used on someone with HIV. Still. HIV is not rare.

Even with a potential half-million exposures a year to sharps, as of December 31, 2013, only 58 confirmed occupational transmissions of HIV and 150 possible transmissions had been reported in the United States. Of these, the vast majority follow hollow needle sticks (88%), where blood drawn from a patient still remains in the bore of the needle which can be reinjected into the unfortunate recipient. If the patient is terminally ill with AIDS, the incidence of transmission by hollow needlestick is increased, meaning the viral load in the blood is important. A solid needle, even a bloody one, presents a very low risk (the risk is, in fact, unknown*.) Getting blood splashed in the eye is a higher risk than a sharp solid (i.e. suture or lancet) needle.

The piercing mouthparts (The mandibles and the maxillae) are solid, akin to a solid sharp. Only the twin tubes (the hypopharynx and the labrum) actually reach blood.

So, given the combined factors of limited survival of HIV on exposed surfaces, the small size of the proboscis (compared to a suturing needle or a lancet, both of which have a higher surface area, yet not known for certain ever to have transmitted HIV), the low probability that the mouth parts have HIV-laden droplets of blood on it, the fact that only the tip of the hypopharynx and labrum encounter blood (the fluid most likely to contain virus) and the extremely low probability that there are enough actual HIV viruses to cause an infection anywhere on the proboscis, I hope this answers your question.

Edited to add: I've focused on HIV here, because it's not known to infect mosquitoes. Mosquitoes are considered by many to be the most dangerous animal in the world. They transmit a significant number of viruses and parasites (like dog heartworm, elephantiasis, dengue fever, equine encephalitis, etc.), but (hypothesized from those diseases that have been studied in mosquitoes) only after being infected by the pathogens themselves. Some die of the infection; some die before the infection spreads through the mosquito's body, some don't harm the mosquito. But transmission is from infected saliva injected into or onto the bite site, not virus particles on the proboscis, as evidenced by the need for adequate incubation time between the bite that takes up the virus to the bite that transmits the virus. If there was sufficient virus on the surface of the proboscis, that incubation time would not be necessary.

Best infection control practices for intradermal, subcutaneous, and intramuscular needle injections
Workbook for Designing, Implementing and Evaluating a Sharps Injury Prevention Program
Occupational HIV infection among health care workers exposed to blood and body fluids in Brazil
Needlesticks: What you must know

  • $\begingroup$ Great answer! The only thing this leaves me wondering is, how does sharing needles lead to virus transmission? I believe it's common among drug users (especially hepatitis I believe?). I guess hypodermic needles can suck small amounts of blood inside the tip? $\endgroup$ Commented Apr 5, 2016 at 20:45
  • 2
    $\begingroup$ Hepatitis B is much more easily transmitted by sharps and needlesticks than HIV. In terms of needle sharing, some of it has to do with withdrawing a small amount of blood into the syringe before injecting the drug, guaranteeing that some blood will be in the syringe shared with the next person, and the fact that it's a hollow-bore needle. $\endgroup$ Commented Apr 5, 2016 at 20:48
  • $\begingroup$ I assume the "withdrawing a small amount of blood" is accidental and related to untrained needle use (e.g. allowing the plunger to move out slightly before withdrawing the needle)? I guess the obvious difference with a mosquito is, that any remaining inside their proboscis will then be sucked into their stomach, whereas with a shared needle it'll be mixed with the next injection. $\endgroup$ Commented Apr 5, 2016 at 21:07
  • 4
    $\begingroup$ I have never injected drugs into myself, but the usual procedure insuring that the needle is in the vein is to withdraw the plunger a bit until you get "venous return" (blood tinging the fluid in the hub if not the barrel of the syringe.) If you don't do this, you might inject the drug into subcutaneous tissue instead of the blood stream, which would be pretty much a waste of the drug. $\endgroup$ Commented Apr 5, 2016 at 21:16
  • $\begingroup$ @user568458 anongoodnurse is, of course, correct, but I'd just add: IV needle sharing involves repeated direct injection into a vein. The direct injection into a vein makes it a much more efficient method for transmitting any bloodborne pathogen than a single prick of the skin. And yes, the transmission efficiency of HBV > HCV > HIV, but when I had my sleep deprived accidental stick with a sharp used on an HIV+ patient, I surely didn't inject it into my vein. $\endgroup$
    – De Novo
    Commented Aug 3, 2018 at 17:38

What you're talking about is called mechanical transmission (at least in veterinary and medical circles; plant epidemiologists call it something different). There's a nice overview by Gray and Banerjee here, although it's a bit old.

The short answer is that blood-feeding insects can mechanically transmit a range of pathogens, including viruses. Mosquitoes specifically are thought not to be great mechanical vectors because of the small quantities of infectious material that could be transferred between feeds. Large biting flies such as tabanids are thought to be much better potential mechanical vectors as they are pool-feeders; they basically chew up an area of skin and then mop up blood from the wound with their mouths. The quantity of blood contaminating a horse fly's mouthparts after feeding is reckoned to be about 10 nanolitres (Foil et al. 1987), which while not much could contain an infectious dose for some viruses reaching high titres in the blood. They also feed frequently (and painfully, which means they are often interrupted).

However, a number of viruses do seem to be transmitted in the field or in a lab setting by mosquitoes, particularly poxviruses (lumpy skin disease virus, myxoma virus and fowlpox virus, for example). Equine infectious anaemia virus (a retrovirus) is also thought to be mechanically transmitted by large biting flies in some regions despite its relative fragility. Some parasitic and bacterial diseases are also mechanically transmitted by blood-feeding insects; Trypanosoma evansi for example (the causative agent of surra) has lost the capacity to be biologically transmitted by tsetse and is solely mechanically transmitted.

(suggestions for improving this answer extremely welcome; I'm fairly new to SE)


You must log in to answer this question.

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