There are two areas, and they are not the same number

When anyone states the area of a Japanese property, the first question is which of these two they mean. In a listing it is almost always the first.

Why the gap exists

The reason is historical and it is the same reason across most of the country. Registered areas largely trace back to land surveys conducted for a nineteenth-century tax reform, carried out with the instruments and methods of the time.

There was also an incentive built into the exercise. Tax was assessed on area, so there was a tendency to report less land than was actually there.

The consequence is land throughout Japan that measures larger than the registry says when someone finally surveys it. Japanese has a word for that surplus, and a matching one for the opposite case where the survey comes out smaller. Both are ordinary findings rather than anomalies.

Parcels have since been split, merged, re-recorded and passed along. The registered figure is best understood as an entry in a ledger, not a description of the ground.

Which land is most affected

Not evenly. Urban subdivisions developed in recent decades were surveyed when they were laid out and their figures are generally reliable.

The discrepancies concentrate in rural land, woodland, farmland and old parcels that have not been split for a long time. That is precisely the category of property overseas buyers are drawn to. The countryside house with the large plot is the one most likely to have an area figure nobody has tested in living memory.

The maps carry the same problem

The Legal Affairs Bureau holds maps showing the shape and position of parcels, but they are not all of one kind.

Some are proper survey-based maps, where shape and position can be relied on. Others descend from those same nineteenth-century drawings. They are useful for understanding which parcel adjoins which, and they are not a source from which dimensions or areas can be read off.

Large parts of Japan are still covered only by the second kind. That a map exists does not establish that the boundary is known.

The contract decides who absorbs the difference

Japanese practice has two ways of handling this, and they produce opposite outcomes.

Buy on the first basis, discover afterwards that the plot is twenty square metres smaller than the registry claimed, and there is normally nothing to reclaim. The contract said what it said.

Which method applies is written in the contract. If you are signing a Japanese contract you cannot read, this is one of the few points worth insisting on having identified before you sign.

When somebody says there is a survey, ask which kind

Several quite different documents get described in English as a survey drawing. The difference between two of them decides whether the boundary is actually settled.

The first assumes the wall is in the right place. If somebody built that wall forty years ago roughly where they thought the line ran, the drawing repeats the same error with great precision.

A confirmed survey requires every adjoining owner to attend and agree. Where the plot fronts a road or a watercourse, the authority attends as well. With several neighbours it takes months, and if one of them has died and the heirs cannot be traced, it does not complete at all.

Who does this work

Confirming land boundaries and reflecting the result in the registry is the province of a specific qualified professional in Japan, distinct from the technician who operates the instruments. The roles are separate, and which specialist you need changes with the situation. For a foreign buyer this is worth knowing simply because "I'll get a surveyor" does not map onto a single Japanese profession.

The markers on the ground

Boundaries are marked physically, using concrete stakes, plastic stakes, metal plates or metal studs set into a road surface or a wall cap. Whether you will find them depends on the type and on the history of the site.

Markers disappear for ordinary reasons. They get pulled out or buried during building work, covered by soil and planting, lost when a drive is resurfaced, or were never set at all. And occasionally they were set, but not in the correct position.

People ask how many of the four corners are usually findable. The average is not a useful number. What matters is when the markers were set and what has happened to the site since. On a recent subdivision they are generally intact. In an area lived in for generations, decades of building and landscaping have removed them a few at a time.

When you or somebody acting for you visits, go and look at all four corners and record what is there. Whether markers can be found is one of the fastest indicators of how settled the land actually is.

From practice

Working in the surveying field in a residential area of Osaka Prefecture, we met this directly. Two neighbours both believed the block wall between them was the boundary. When the land was actually surveyed, the wall stood somewhere else entirely.

Neither of them was being dishonest. Each had inherited the understanding from the previous owner, and it had passed down unchallenged for decades. The registered area and the surveyed area differed substantially, and that was unremarkable too.

This is not an accident that befalls unlucky properties. On land nobody has dealt with properly for a long time, it is simply what is there.

Districts where the maps and the ground diverge

Japan has areas where the official maps and the physical ground disagree across an entire district rather than on one parcel.

The shapes on the map do not match the shapes on the ground. The registered areas do not sum to the area of the block. Adjoining owners hold different views about where the lines run. An attempt to confirm one boundary draws in far more parties than expected. Causes vary and include old development and grading history, records that were not carried forward properly, and reorganisation after disasters.

Nothing in a listing tells you the land is in such an area. Not the price, the photographs or the floor plan. It emerges when somebody looks at the map at the Legal Affairs Bureau, searches for markers on site, and asks the authority.

This is not a reason never to buy there. It is a reason to budget, in time and money, on the basis that establishing the boundaries will be a project. The same issue will be waiting when you come to sell.

Apartments have a different area problem

Buying a unit rather than land does not remove the question, it changes it.

A unit can be measured to the centre of its walls or to the inside face of them. Sales material generally uses the first. The registry records the second. The registry figure is always the smaller of the two.

This matters beyond tidiness. Several Japanese tax reliefs carry floor area requirements, and some of them, including the income tax credit for housing loans, are assessed on the registry figure rather than the brochure figure. A unit advertised as comfortably above a threshold can sit below it in the registry, and the relief goes with it.

Two further points. The thresholds are not the same across different reliefs, neither is the way the area is measured, and both change with tax reform. One of them was lowered with effect from April 2026, which means a good deal of the guidance currently online states the previous figure. Check the registry area, and check the threshold that applies now to the relief you are relying on. Costs and reliefs at purchase are covered further in what a cheap Japanese house actually costs.

What goes wrong when the area is wrong

Look at the resale row. What you skip when buying arrives again when you sell, and by then you are the one being asked to resolve it.

Boundary questions also decide who owns the wall between two plots at different levels, which is a larger problem than it sounds. That is dealt with in the wall holding up the land is part of what you are buying.

Seven things to establish

1. Is the area in the listing the registered figure or a surveyed one?
2. Does a confirmed survey exist, and what date does it carry?
3. Can boundary markers be found on site, at all four corners?
4. Is there a boundary confirmation document or an encroachment memorandum with the neighbours?
5. Is the contract priced on the registered area or the surveyed area, and if the latter, who pays for the survey?
6. For an apartment, what is the registered area of the unit?
7. Is the land in a district where the maps and the ground are known to diverge?

On the second, a confirmed survey decades old may no longer describe the situation. The date matters as much as the existence. The rest of the pre-purchase checks, and which office answers each one, are set out in what you must verify yourself before buying.

Land does not move

Area is a dull subject and it loses easily to price and photographs. But the land stays where it is, and on the day you buy it you inherit whatever history it carries.

Buying land whose boundaries were never confirmed, and leaving them unconfirmed, is entirely possible in Japan and happens constantly. Doing that knowingly and doing it unknowingly are different transactions.