Candiru (Vandellia cirrhosa), also known as cañero, toothpick fish, or vampire fish, is a species of parasitic freshwater catfish in the family Trichomycteridae native to the Amazon basin where it is found in the countries of Bolivia, Brazil, Colombia, Ecuador, and Peru. The name "candiru" is also used to refer to the entire genus Vandellia.

This species is known for an alleged tendency to invade and parasitize the human urethra and other bodily openings; however, despite ethnological reports dating back to the late 19th century, the first documented case of the removal of a candiru from a human urethra did not occur until 1997, and even that incident has remained a matter of controversy.

Vandellia cirrhosa is a small, freshwater catfish. Members of the genus Vandellia can reach up to 17 cm (7 in) in standard length, but some others can grow to around 40 cm (16 in). The fish has an elongated body with an anterior dorsal fin and pelvic fin, and an anal fin slightly larger than the dorsal fin. The caudal fin is fairly small with a truncated shape. Each has a rather small head and a belly that can appear distended, especially after a large blood meal. The body is translucent, making it quite difficult to spot in the turbid waters of its home. Blood is often visible through the translucent body after feeding. The fish has small eyes with yellow irises. There are short sensory barbels around the head, together with short, backward pointing spines on the gill covers. These spines have been described as popping out in an umbrella-like fashion, which could be used to help lodge the fish into its host.

All members of the subfamily Vandelliinae share the traits of blood parasitism, with parasitism in general being a shared ancestral trait of all members of Trichomycteridae. They have individual claw-like teeth for this purpose. Their bodies are very small and elongated to easily slip into the gills of host fish.

Candiru inhabits the Amazon and Orinoco basins of lowland Amazonia.

This fish lives in shallow water with muddy, sandy, or rocky bottoms. It can be found in riffles. Its distribution is patchy and it does not seem to move very far from its spawning sites.

One location that Vandellia cirrhosa is specifically known to inhabit is the Purus River of Brazil. This location is hard to study due to its geographic isolation, something that is common among the habitats of candiru.

Candiru is hematophagous and parasitizes the gills of larger Amazonian fish, especially catfish of the family Pimelodidae (Siluriformes) and members of the family Characidae. However, it has been known to parasitize many species in the same location, suggesting that its feeding habit is based more on availability than species preference. Vandellinae is one of only two groups of jawed vertebrates that exclusively feed on blood.

The feeding mechanism of the candiru was not understood until fairly recently, but many theories had been proposed before. Some suggested that it uses its sharp teeth to latch onto an artery or vein and stays attached to the host until it has ingested enough blood. It then lets go of the host and continues swimming. Others suggested that it extracts blood from its hosts by latching onto the gill filaments, which bleed freely into the alimentary tract of the candiru. In one experiment involving Vandellia cirrhosa, no evidence of gill damage was found on the fish hosts. Thus, the experimenter suggested a hypothetical blood-pumping mechanism in which the candiru is able to quickly ingest large amounts of blood without permanently damaging the host.

Modern experiments have shown that the candiru feeds by approaching a host fish and swimming alongside it until close to the gill cover. It then attempts to penetrate the gill chamber by forcing itself underneath. The candiru has also been seen entering the host through its mouth, though this behavior seems to be rare. While latched onto the host fish’s gill chamber with sharp teeth, the candiru goes limp and quickly swells up with blood. The gut of this species is a straight tube with loosely-spaced fibers lining the walls of the connective tissue, most likely facilitating the swelling of the belly that is associated with the candiru. The lack of any protrusible jaw supports the theory that this species does not suck blood, but rather bites into a blood vessel and ingests the fluid that flows freely into the mouth. Because the candiru relies on the blood pressure of the host to ingest blood from the ventral or dorsal arteries, host fish must be selected by size.

The time taken to get its fill of blood depends on the size of the candiru and whether it has attached to a large or small blood vessel. Because of the small size of Vandellia cirrhosa, it generally seems to take no more than two minutes to ingest the required amount of blood from the host fish. This short duration is theoretically beneficial to the candiru because it is only vulnerable to predators for a short period of time.

In most cases, the host fish do not seem to be badly wounded by this process. There is generally no observable damage to the gill filaments. However, relatively deep crescent or elliptical-shaped wounds with coagulated blood inside can be found beneath the gill cover.

When starving, the candiru may resort to entering unusual orifices such as the nostril of a host fish. This behavior may relate to reported cases of these fish penetrating human orifices such as the urethra.

Although lurid anecdotes of attacks on humans abound, only one somewhat questionable case has evidence behind it, and some alleged traits of the fish have been discredited as myth or superstition. It is likely that, while the fish's spines can cause physical trauma, it merely poses as much danger of actually entering a human as any other fish of its size.

The earliest published report of candiru attacking a human host comes from German biologist C. F. P. von Martius in 1829, who never actually observed it, but rather was told about it by an interpreter relaying the speech of the native people of the area, who reported that men would tie ligatures around their penises while going into the river to prevent this from happening. Other sources also suggest that other tribes in the area used various forms of protective coverings for their genitals while bathing, though it was also suggested that these were to prevent bites from piranha. Martius also speculated that the fish were attracted by the "odor" of urine. Later experimental evidence has shown this to be false, as the fish actually hunt by sight and have no attraction to urine at all.

Another report, from French naturalist Francis de Castelnau in 1855, relates an allegation by local Araguay fisherman, saying that it is dangerous to urinate in the river as the fish "springs out of the water and penetrates into the urethra by ascending the length of the liquid column." While Castelnau himself dismissed this claim as "absolutely preposterous," and the fluid mechanics of such a maneuver defy the laws of physics, it remains one of the more stubborn myths about the candiru. It has been suggested this claim evolved out of the real observation that certain species of fish in the Amazon will gather at the surface near the point where a urine stream enters, having been attracted by the noise and agitation of the water.

In 1836, Eduard Poeppig documented a statement by a local physician in Pará, known only as Dr. Lacerda, who offered an eyewitness account of a case where a candiru had entered a human orifice. However, it was lodged in a native woman's vagina, rather than a male urethra. He relates that the fish was extracted after external and internal application of the juice from a Xagua plant (believed to be a name for Genipa americana). Another account was documented by biologist George A. Boulenger from a Brazilian physician, named Dr. Bach, who had examined a man and several boys whose penises had been amputated. Bach believed this was a remedy performed because of parasitism by candiru, but he was merely speculating, as he did not speak his patients' language. American biologist Eugene Willis Gudger noted that the area which the patients were from did not have candiru in its rivers, and suggested the amputations were much more likely the result of having been attacked by piranha.

In 1891, naturalist Paul Le Cointe provides a rare first-hand account of a candiru entering a human body, and like Lacerda's account, it involved the fish being lodged in the vaginal canal, not the urethra. Le Cointe supposedly removed the fish himself, by pushing it forward to disengage the spines, turning it around and removing it head-first.

However, the veracity of both Le Cointe's and Poeppig's accounts are questionable, due to a trend of Europeans from various careers residing in Brazil including scientists, "explorers, medical men, and missionaries" regularly using exaggerated accounts of native people to advance their economic and social status through writing and building rapport with others with similar positions.

Gudger, in 1930, noted there have been several other cases reported wherein the fish was said to have entered the vaginal canal, but not a single case of a candiru entering the anus was ever documented. According to Gudger, this lends credence to the unlikelihood of the fish entering the male urethra, based on the comparatively small opening that would accommodate only the most immature members of the species.

To date, there is only one documented case of a candiru entering a human urethra, which took place in Itacoatiara, Brazil, in 1997. In this incident, the victim (a 23-year-old man named Silvio Barbossa, also known as "F.B.C.") claimed a candiru "jumped" from the water into his urethra as he urinated while thigh-deep in a river. After traveling to Manaus on October 28, 1997, the victim underwent a two-hour urological surgery by Dr. Anoar Samad to remove the fish from his body.

In 1999, American marine biologist Stephen Spotte traveled to Brazil to investigate this particular incident in detail. He recounts the events of his investigation in his book Candiru: Life and Legend of the Bloodsucking Catfishes. Spotte met Dr. Samad in person and interviewed him at his practice and home. Samad gave him photos, the original VHS tape of the cystoscopy procedure, and the actual fish's body preserved in formalin as his donation to the National Institute of Amazonian Research. Spotte and his colleague Paulo Petry took these materials and examined them at the institute, comparing them with Samad's formal paper. While Spotte did not overtly express any conclusions as to the veracity of the incident, he did remark on several observations that were suspicious about the claims of the patient and/or Samad himself.

• According to Samad, the patient claimed "the fish had darted out of the water, up the urine stream, and into his urethra." While this is the most popularly known legendary trait of the candiru, according to Spotte it has been known conclusively to be a myth for more than a century, as it is impossible because of simple fluid physics.
• The documentation and specimen provided indicate a fish that was 133.5 mm in length and had a head with a diameter of 11.5 mm. This would have required significant force to pry the urethra open to this extent. The candiru has no appendages or other apparatus that would have been necessary to accomplish this, and if it were leaping out of the water as the patient claimed, it would not have had sufficient leverage to force its way inside.
• Samad's paper claims the fish must have been attracted by the urine. This belief about the fish has been held for centuries, but was discredited in 2001. While this was merely speculation on Samad's part based on the prevailing scientific knowledge at the time, it somewhat erodes the patient's story by eliminating the motivation for the fish to have attacked him in the first place.
• Samad claimed the fish had "chewed" its way through the ventral wall of the urethra into the patient's scrotum. Spotte notes that the candiru does not possess the right teeth or strong enough dentition to have been capable of this. Additionally, the fish would most likely have died before it could have chewed even a somewhat large part of what was needed to reach it.
• Samad claimed he had to snip the candiru's grasping spikes off in order to extract it, yet the specimen provided had all its spikes intact.
• The cystoscopy video depicts traveling into a tubular space (presumed to be the patient's urethra) containing the fish's carcass and then pulling it out backwards through the urethral opening, something that would have been almost impossible with the fish's spikes intact.

When subsequently interviewed, Spotte stated that even if a person were to urinate while "submerged in a stream where candiru live", the odds of that person being attacked by candiru are "(a)bout the same as being struck by lightning while simultaneously being eaten by a shark."

Candiru belong to the family Trichomycteridae, which is monophyletic and contains 207 species. The taxonomic placement of this fish has long been debated, with the first proposed phylogenetic relationships of Trichomycteridae being proposed by Jonathan N. Baskin in 1973. Most proposed phylogenies have relied on morphological data, often placing Vandellinae and Stegophilinae as sister taxa among the subfamilies. A study conducted by Luis Fernández and Scott A. Schaefer, published in 2009, used DNA sequence data to create the first comprehensive treatment of phylogenetic relationships of trichomycterid catfish. Relationships among Vandelliinae were strongly supported, and Vandellia was found to be the sister group of Plectrochilus. The results of Fernández and Schaefer were fully congruent with previous statements based on morphological data. Nonetheless, the taxonomy of the Vandellia genus is still incomplete and hindered by the fact that several species within the genus have often been attributed the same name.

Vandellia cirrhosa was discovered in the early 1800s by Alexandre Rodrigues Ferreira, a Native Brazilian scholar studying under the Italian naturalist Professor Domingos Vandelli, of which the fish would be named after. The Spanish name cañero is a derivative of carnero, meaning flesh-eater.

One of the most well known scientific mentions of the candiru appeared in The American Journal of Surgery published in 1930, summarizing the supposedly centuries old tale of a fish that penetrates the urethras of nude bathers in the Amazon.

The remote habitat of the candiru, as well as the indigenous cultural customs surrounding its location, makes it difficult to study. The number of Vandellia cirrhosa in the wild is unknown, but there are no conservation efforts in place to protect these fish.

One of its main habitats, the Purus River, is currently the main source of fish for human consumption in the most populous city of the Central Amazon, Manaus. This creates a huge pressure on fish stocks, which may be indirectly affecting the candiru by depleting its population of potential host fish.