1,388 markets, 202,039 stalls
Korea officially recognises 1,388 traditional markets, and they contain 202,039 stalls between them. That is an average of 146 stalls per market — though as we will see, the average is one of the least informative numbers in this dataset.
These are the markets local governments formally register, which means the dataset is a census rather than a sample. Every market has a stall count except five, every one lists the categories of goods it sells, and every one records whether it has a public toilet and parking. For Korean open data, this is unusually complete.
It is worth stating one methodological point up front, because it differs from the other datasets we write about. In Korea’s public toilet data, a blank field is stored as “no”, so we can only ever report registration rates. Here, blanks are omitted rather than converted — and no market has a blank. When this data says a market has no parking, it means the survey recorded no parking.
Markets and stalls by province
| Province | Markets | Stalls | Mean | Median |
|---|---|---|---|---|
| Seoul | 189 | 43,749 | 231 | 87 |
| Busan | 160 | 26,638 | 166 | 90 |
| Gyeongnam | 155 | 17,427 | 115 | 78 |
| Gyeonggi | 150 | 23,040 | 154 | 108 |
| Gyeongbuk | 136 | 15,849 | 117 | 68 |
| Jeonnam & Gwangju | 118 | 9,695 | 82 | 56 |
| Daegu | 109 | 16,947 | 155 | 85 |
| Gangwon | 61 | 7,566 | 124 | 98 |
| Chungnam | 57 | 7,420 | 133 | 94 |
| Jeonbuk | 57 | 4,718 | 83 | 54 |
| Chungbuk | 55 | 6,055 | 110 | 72 |
| Incheon | 44 | 11,504 | 261 | 122 |
| Ulsan | 42 | 4,161 | 101 | 80 |
| Daejeon | 28 | 2,935 | 105 | 73 |
| Jeju | 23 | 3,752 | 163 | 78 |
| Sejong | 4 | 583 | 146 | 87 |
| All markets | 1,388 | 202,039 | 146 | 83 |
Seoul has the most markets (189) and the most stalls (43,749), which is what you would expect. Sejong, a purpose-built administrative city with no old town to inherit, has four.
The gap between the mean and median columns is the thing to notice. Nationally the mean is 146 and the median is 83 — a typical Korean market is barely half the size the average implies, because a handful of enormous wholesale complexes drag the mean upward. Seoul is the extreme case: mean 231, median 87.
416 markets still run on a five-day cycle
The dataset records how often each market opens, and this is where it stops being an inventory and starts being interesting. 972 markets (70.0%) open daily. The other 416 (30.0%) open on a five-day cycle — the traditional 오일장 (*ojang*), a rotating rural market system that predates modern retail by centuries.
| Opening cycle | Markets | Share |
|---|---|---|
| Daily | 972 | 70.0% |
| Days ending 2 and 7 | 91 | 6.6% |
| Days ending 5 and 10 | 90 | 6.5% |
| Days ending 4 and 9 | 84 | 6.1% |
| Days ending 3 and 8 | 82 | 5.9% |
| Days ending 1 and 6 | 68 | 4.9% |
| Days ending 2, 4, 7 and 9 | 1 | 0.1% |
| All markets | 1,388 | 100% |
The logic is elegant and worth understanding before you plan a trip around one. A market listed as “2 and 7” opens on every date whose last digit is 2 or 7 — the 2nd, 7th, 12th, 17th, 22nd and 27th of the month. Five days later, always. Neighbouring towns take different digits so that traders can work a circuit, which is why the five variants are distributed so evenly: 68 to 91 markets each, none dominant.
That even spread is not an accident of the data. It is the surviving imprint of a rotation designed so no two nearby markets competed on the same day, and it is still legible in a government spreadsheet in 2026.
If you want to visit one, check the date arithmetic before you travel. A five-day market on the wrong day is an empty car park. This is the single most common way visitors miss a Korean market — the listing looks like a permanent place because it is a permanent place, but it is only alive six days a month.
What Korea calls a “market”
One reason the size range is so wide — from four-stall village markets to 남대문시장’s 5,111 — is that the word covers several genuinely different things, and the dataset does not sort them for you. Reading through it, four broad kinds emerge.
- Neighbourhood retail markets, the great majority: a covered street or arcade of food stalls, clothing shops and cheap restaurants serving the surrounding blocks.
- Five-day rural markets, the 오일장: periodic markets that assemble in a small town on fixed dates and disperse again, often the main shopping event for a farming district.
- Historic mega-markets, a handful: 남대문, 동대문 and their peers, which function as wholesale districts and tourist destinations at once.
- Specialist and wholesale complexes: tools, industrial goods, textiles, fish. These carry the largest stall counts and the least resemblance to what a visitor pictures as a market.
The single field that lets you tell these apart at a glance is the merchandise category. Every market lists what it sells, from the single-word “축산물” (livestock products) of a specialist meat market to the eight-category sprawl of a general market carrying produce, meat, fish, processed food, clothing, footwear, household goods and prepared meals. A long list means a general market; a short one usually means a specialist.
Need this near you right now? Open the app.
Open Bballi BballiToilets and parking
1,314 markets (94.7%) have a public toilet and 1,154 (83.1%) have parking. Because blanks are omitted rather than defaulted in this dataset, those are genuine possession rates rather than the paperwork proxies we have to use elsewhere.
Toilet provision is close to universal and does not vary much: Jeonbuk, Ulsan, Jeju and Sejong report 100%, and the lowest is Busan at 90.6%, with Seoul at 91.0%. A twelve-point spread across the whole country is a strong result.
Parking is a different story entirely, and it runs the opposite way to what you would guess. Seoul is last at 65.1%, against a national figure of 83.1% and a perfect 100% in Jeonbuk and Sejong. Korea’s biggest cities have its worst market parking — we look at why that is, and what it means for visiting, in a separate piece.
What this data cannot tell you
Two limits are worth flagging, both of which we ran into while counting.
We do not know how current this is. The national standard includes a data-reference-date field, but our copy of the dataset carries no date on any of the 1,388 records — the field was not carried through when it was loaded. Treat stall counts as approximate and possibly a year or more old. Markets shrink and close, and this data will show them open.
The other is that a stall count is not a measure of how busy a market feels. 남대문시장 in Seoul records 5,111 stalls, the largest in the country, and behaves like a district rather than a market. Meanwhile a 60-stall five-day market in a small town can be the busiest place for miles on the day it opens. The number measures floorspace, not life.
What the data does support is planning: where markets are, how big they are, when they open, and whether you can park. For a visitor those four facts cover most of what you need, and Korea publishes all four for every registered market in the country — which is more than most countries manage.
Sources and method
- Source: 전국전통시장표준데이터 — National Traditional Market Standard Data, Small Enterprise and Market Service (소상공인시장진흥공단), via the spatial edition published on VWorld.
- Counts taken from our catalogue on 6 August 2026: 1,388 markets, 202,039 stalls across the 1,383 records carrying a stall count.
- Toilet and parking figures are possession rates, not registration rates: the loader omits missing values rather than converting them to “no”, and no record was missing either field.
- Publication date unknown — our copy carries no reference date on any record.