Nearly every supermarket banana is a copy from one small group of plants
Most supermarkets in the world sell the same kind of banana, called Cavendish. It has three sets of chromosomes where most plants have two, so it is sterile and cannot be grown from seed.
So growers copy it. A new plant starts either as a shoot taken from the base of an existing one, called a sucker, or as a piece of plant tissue multiplied in a laboratory. Either way, the new plant is genetically a copy of its parent. You may see names on a box such as Grande Naine, Williams or Valery. These are not unrelated bananas. They are varieties within the Cavendish group : copies that turned out slightly different, for example in height, and were kept because they grew well. Strictly, then, Cavendish is a small group of closely related copies rather than one single plant.
That is the first half of the answer. A banana bought in Seoul, Paris or São Paulo tastes so alike because, for practical purposes, it comes from the same small group of copies.
Most of the world's bananas are not this banana
This is the part that surprises most people. Cavendish is not the world's banana. It is the world's exported banana.
In a review of the years up to 2002, the UN Food and Agriculture Organization (FAO) put Cavendish at about 43 percent of the world's banana production. The same review found that only about 11 percent of the world's harvest was traded between countries at all. ProMusa, an online reference on bananas, gives a later estimate for 2018. It puts Cavendish at 57 percent of world production that year, with Cavendish exports equal to about 16 percent of world production. FAO's current overview describes Cavendish as a little less than half of what is grown.
The estimates differ by period and method, but they agree on the shape. Cavendish is roughly half of what is grown. Most bananas of every kind are eaten in the country that grows them, and many of those are kinds a supermarket shopper never sees.
Across borders, the picture flips. In its account of the years up to 2002, FAO described virtually all exported bananas as Cavendish, and a 2017 paper in Nature Communications uses the same words, 'virtually all the export trade'. Figures of 95 or 99 percent are often quoted, but they come from secondary sources, not from a published count. 'Nearly all' is as precise as the evidence allows.

Picked green, shipped cold, ripened on a timetable
Genes are only half of the sameness. Export bananas are cut while still green and hard. They travel in refrigerated containers at about 13 degrees Celsius, cold enough to hold ripening back. In the importing country they go into sealed ripening rooms, usually kept somewhere between 15 and 20 degrees ; one common schedule holds them at around 17 degrees. There a small dose of ethylene, a gas that ripening fruit gives off naturally, starts all of them ripening at once. On that schedule they leave for the shops three or four days later, at the colour the buyer ordered.
So the bananas on the shelf are not only copies from the same small group of plants. They have also made much the same journey, at similar temperatures, on a similar schedule. The trade makes ripening uniform on purpose, because only a fruit that ripens predictably can be sold across an ocean.
Alike is not identical, though. How mature the fruit was when it was cut still shapes its sugars, acids and aroma, so two Cavendish bananas can taste different. What the trade delivers is a banana that is unusually consistent, not one that is the same every time.

A fungus has already driven out one export banana
For the first half of the twentieth century, the banana in Western shops was a different one, called Gros Michel. Traders liked it because it produced large bunches and had a peel that survived long journeys. It too was grown from copies.
Then a soil fungus spread through the plantations. The disease it causes is called Fusarium wilt, or Panama disease. Fungicides cannot control it and fumigation cannot clear it from the soil. It travels in infected planting material, soil and water, and an infected shoot can look healthy when it is planted somewhere new. Because Gros Michel was grown from copies, a weakness in one plant was a weakness shared across the plantations. Growers kept abandoning infected land for new land. In its account of Central America, ProMusa puts the abandoned plantations at more than 40,000 hectares.
The way out was a different banana. Cavendish resisted the strain that was killing Gros Michel, now called race 1. In the late 1950s, under pressure from cheaper bananas from Ecuador, Central American growers replanted their infested soils with Cavendish. That is how the trade came to sell the banana it sells today.
A new strain of the old banana fungus kills Cavendish
Cavendish plants in Taiwan were already falling sick by 1967. At the end of the 1980s, scientists studying Taiwanese samples identified the cause as a new strain of the same fungus, one that does kill Cavendish. It is called Tropical Race 4, or TR4. It spread through Asia, reached Africa in 2013 and has since been reported in more than 20 countries, by counts that change often.
The map changes from year to year. The first confirmed report of TR4 in Nepal appeared in the journal Plant Disease, online in June 2026 and then in its September 2026 issue. Ecuador, the world's largest banana exporter, confirmed its first case in December 2025, after a suspected outbreak in September, on a farm in El Oro province near the border with Peru. In July 2026 its plant health agency, Agrocalidad, confirmed a second case on a neighbouring farm.
The cases reported so far are a warning, not a collapse. Agrocalidad says it has isolated the area and tightened controls in the province and at the southern border. In an update on 3 November 2026 it described containment in El Oro as still under way, without giving a current count of affected farms. The number of cases may have changed since.

Why there is no ready replacement this time
Last time, a resistant banana already existed when the old one failed. This time the candidates are less certain.
One is Formosana, a Cavendish variant selected in Taiwan. In Taiwanese field trials it fell sick less often than the Cavendish it was compared with. In a three-year trial in Australia's Northern Territory, though, it proved as susceptible to TR4 as the ordinary Cavendish variety Williams. How well it holds up depends on where it is tested.
Another is QCAV-4, developed by Queensland University of Technology, which added a resistance gene from a wild banana to a Grande Naine Cavendish. In six years of field trials in the Northern Territory it proved highly resistant. In 2024 Australian regulators approved it for both growing and eating, the first approval of a genetically modified banana anywhere. At that time the university said it had no immediate plans to grow it commercially, and that it would be kept as a safety net in case the disease spread through Australian farms. It is also still a Cavendish : one plant, copied, with the resistance gene added, along with a marker gene used in the laboratory to pick out the modified plants.

The same taste and the same weakness are one fact
Put the two halves together and the answer to the question is also a warning. Bananas taste so alike because the trade grows one small group of copies and handles them much the same way from field to shop. That same sameness is why a fungus that can kill one plant can sweep through a whole plantation.
None of this means bananas are about to disappear. Most of the world's bananas are other kinds, eaten where they are grown, and Gros Michel is still grown in places today. What it does mean is that the banana in your kitchen is a choice the trade made once, after a fungus forced its hand.









