21,179 Accessible Public Toilets in Korea

What 51,228 records actually say

Korea publishes its public toilets as open government data. We aggregate that data to run this app, and as of August 2026 our catalogue holds 51,228 records. Of those, 21,179 — 41.3% — are marked as having accessible facilities for wheelchair users.

That single number is easy to misread, so here is the important part first: 41.3% is a registration rate, not a presence rate. It tells you how often a local government wrote the information down. It does not tell you that the remaining 58.7% of toilets lack accessible facilities. Some of them almost certainly have them and simply never had the field filled in.

We think that distinction is the actual story. Once you split the same dataset by province, the gap between the best and worst record-keeping is nearly fourfold — and it does not track wealth, population, or how many toilets a province has.

Accessible-toilet registration by province

Sejong records accessible facilities for 69.8% of its 536 toilets. Jeonbuk records them for 17.8% of its 3,329. That is a 3.9× spread between two provinces reporting into the same national standard, using the same fields.

Volume does not explain it. Gyeonggi has by far the most toilets on file — 12,245, roughly a quarter of the whole country — and still manages a 50.7% registration rate, above the national average. Gyeongnam has the second-most (5,732) and sits near the bottom at 25.6%. Two provinces of similar size, a two-fold difference in how completely they describe what they have.

Why “not recorded” is not “not there”

This is worth being precise about, because it changes what you can honestly say about the data. In the underlying schema, the accessible-facility field is a plain yes/no column that cannot be left empty. When a municipality submits a record without that detail, it does not arrive as “unknown” — it arrives as “no”.

The consequence: “no” and “never filled in” are stored identically. There is no way, from the data alone, to tell a toilet that genuinely has no accessible facilities from one whose paperwork was incomplete. So we only ever count the “yes” values, and we call the result a registration rate.

That is why you will not find a sentence like “58.7% of Korean public toilets are not accessible” anywhere on this site. It would be a much punchier headline. It would also be unsupported by the data we actually have, and it would be unfair to the provinces at the bottom of the table, whose real problem may be a reporting workflow rather than a building.

Read the other way round, though, the table is meaningful. 21,179 toilets are positively documented as accessible. That is a floor, not a ceiling — the true number is higher, and nobody knows by how much.

Seoul is missing from this table, and that is not a bug

If you scanned the table looking for the capital, it is not there. The national catalogue contains exactly one Seoul record. Seoul’s public toilets reach this app through a separate, separately maintained dataset of 4,655 records.

In that Seoul dataset, 2,534 of 4,655 toilets (54.4%) are recorded as having accessible facilities — comfortably above the 41.3% national figure. It is tempting to put that number in the table as a 16th row. We deliberately have not, and the reason is the point of this whole article: the two datasets are collected differently. Seoul’s is curated location by location; the national one is submitted in bulk by hundreds of local authorities. A higher registration rate in a hand-checked dataset says something about the checking, not about the toilets.

Combining them into a single ranking would produce a clean-looking chart and a false comparison. Two numbers can both be correct and still not belong in the same column.

How current these records are

Open data ages, and it ages unevenly. Of the 51,228 records, 43,786 (85.5%) carry a 2026 timestamp, and 42,005 (82.0%) were refreshed in July 2026. At the other end, 467 records (0.9%) have not been touched since 31 December 2021 — well over four years ago.

One honest caveat about those dates: they record when our copy was last refreshed from the public source, not when a local official last visited the toilet and confirmed the facilities. A record can be freshly downloaded and still describe a building as it was years ago.

So treat everything above as a description of the paperwork, which is what open data is. It is genuinely useful for finding a toilet, and genuinely unreliable as a guarantee about what you will find when you arrive.

If you need an accessible toilet in Korea today

• Treat a missing accessibility label as unknown, not as “no”. Especially in Jeonbuk, Gyeongnam and Gyeongbuk, where fewer than a third of records carry the detail, plenty of unlabelled toilets are in fact equipped.
• Subway stations are the most consistent bet. Station facilities are built to a national specification rather than left to individual operators, so they vary far less than street-level public toilets.
• Large, recently built public buildings are the next best. City halls, community service centres, libraries and department stores are covered by accessibility requirements that older standalone park toilets predate.
• Where the data is thin, call ahead. A phone number in a listing is often more reliable than a checkbox in a spreadsheet.

And if you find a toilet whose listing is wrong — accessible when the record says nothing, or blocked when the record says yes — reporting it in the app is the fastest way that information reaches the next person. Community corrections are the only part of this dataset that updates in days rather than years.

Sources and method

The figures above come from the toilet catalogue that powers Bballi Bballi, counted directly rather than estimated. 50,910 of the 51,228 records are public-standard toilet data; the remaining 318 are subway-station toilets. Within the public-standard group, 31,529 records come straight from the national standard dataset published by the Ministry of the Interior and Safety, and the rest arrive from individual provincial and municipal open-data feeds that follow the same schema.

• Primary source: 전국공중화장실표준데이터 — Nationwide Public Toilet Standard Data, Ministry of the Interior and Safety, published on Korea’s public data portal.
• Seoul’s 4,655 records come from the Seoul Open Data Plaza, released under the Korea Open Government License.
• Counting method: a record is counted as accessible only where the accessibility field is explicitly marked yes. Percentages are that count divided by all records in the province.
• A full breakdown of every source we use, and the licence each one carries, is on our /data-sources page.

Figures were counted on 6 August 2026. If you want to check them, the underlying datasets are public and free — that is the whole point of them.