How we work out every landed figure

Where the data comes from

• Private residential transactions — the Urban Redevelopment Authority (URA), through its data service. URA publishes a ROLLING FIVE-YEAR WINDOW, so a sale drops out of our copy when it drops out of theirs. These pages measure the current market; they are not a history.

We mirror the published data. When URA restates or withdraws a record, our copy changes with it.

Why there is no “today's dollars” column here

Our HDB pages restate older prices to today's market level, because those rankings reach back to 1990. Nothing here does. Every window on this page is 24 months or shorter, and over that span the price index moves a median by close to nothing — a restated column would be noise dressed up as rigour. Every figure below is the sum that was actually transacted.

If a longer-horizon ranking is ever added, it will be restated with the URA Property Price Index for its own segment — non-landed for condos, landed for landed — never with the HDB Resale Price Index, and never with one index across both.

Rules that apply to every ranking on this page

• Collective (en-bloc) sales are excluded. One URA record can carry an entire development — Loyang Valley's collective sale went through as 362 units for S$880 million in a single record. Every aggregate here counts only records covering one unit, or a single bulk deal would top the trading tables outright.
• Units sold, not records. Where a table counts homes changing hands it sums the unit count on each record rather than counting rows.
• Grouped by estate AND district. URA reuses estate names across districts, so an estate is identified by both. Two estates sharing a name are ranked separately rather than pooled.
• Windows anchor on the data, never on the clock. Every window is measured back from the most recent month URA has published — Aug 2026 as this page was built — not from today. URA's feed lags by weeks, so a clock anchor would quietly shorten every window for part of each month.
• Ties are broken on name and district, so the same dataset always produces the same order.
• Top 100. Each ranking shows its leading 100 entries; the hub cards show the leading 5.

Landed figures are LAND rates

URA records the plot size for a landed house, not its built-up floor area. Every per-square-foot figure on the landed pages is therefore price divided by LAND area — which is also how landed deals are quoted between agents and buyers.

That makes these numbers incomparable with the psf on our condo pages, which is price over floor area. The two are different measurements that happen to share a unit. We never place them side by side.

For the same reason STRATA landed — cluster housing, where the owner holds a share of the development rather than the plot under the house — is excluded from every landed table. URA records a strata area for it, so its price per square foot is not a land rate: measured islandwide in mid-2026, strata terraces recorded a median of roughly 290 sqm at about S$930 per square foot, against roughly 177 sqm and S$1,900 for terraces on their own plots. Those are two different measurements, and pooling them filled the ascending “most affordable” list with developments that ranked low because of how they are measured rather than what they cost.

Every ranking, its window and its threshold

A median over one or two sales is one household's price wearing a statistic's clothes. Each ranking states the minimum sample an entry must reach before it is eligible, and anything below it is left out rather than shown with a caveat.

Priciest landed estates

Landed estates with the highest median land rate per square foot of plot.

• Window: Landed resale transactions over Sep 2024 – Aug 2026.
• Minimum sample: At least 5 transactions in the window.
• Rows shown: Top 100.
• Read with: A land rate is the price divided by the LAND area — the plot — which is how landed deals are quoted. URA records plot size for landed houses rather than built-up floor area, so the ordinary floor-area psf you see on condos cannot be worked out here, and these two numbers are not comparable. Strata landed — cluster housing, where you own a share of the development rather than the plot under your house — is excluded throughout: URA records a strata area for it, so its price per square foot is not a land rate at all.

Most affordable landed estates

Landed estates with the lowest median land rate per square foot of plot.

• Window: Landed resale transactions over Sep 2024 – Aug 2026.
• Minimum sample: At least 5 transactions in the window.
• Rows shown: Top 100.
• Read with: A land rate is the price divided by the LAND area — the plot — which is how landed deals are quoted. A low rate is usually low for a reason, and the commonest one is a short remaining lease: several of the cheapest estates here are 99-year leasehold from the 1960s and 70s with a few decades left to run. Check the tenure before reading a low rate as a bargain. Strata landed (cluster housing) is excluded from every landed table — URA records a strata area for it, not a plot, so including it would fill this list with developments that rank low because of how they are measured rather than what they cost.

Most traded landed estates

Estates where the most landed homes changed hands on the resale market.

• Window: Landed resale transactions over Sep 2024 – Aug 2026.
• Minimum sample: None — this table counts published transactions rather than averaging them.
• Rows shown: Top 100.
• Read with: A collective (en-bloc) sale transfers many houses in one record, so those are excluded — otherwise a single deal would sit at the top of this list. What is counted is individual houses changing hands.

Known limitations

• An estate is not a house. A landed estate holds terraces, semi-detached and detached houses on very different plots. An estate median is a median across that whole range, not the value of any one house in it.
• A low land rate usually has a reason. Short remaining lease and strata title both depress it. Tenure is not folded into the ranking — read it alongside.
• No detail pages, by design. Landed rows are named but not linked: there is no per-estate report behind them, and a link to a page that does not exist is worse than plain text.
• Five years is the whole record. URA's rolling window is the limit of what we can show. A ranking here cannot reach further back, and we do not keep an archive of sales that have aged out.

What these numbers are not

They are not a valuation. A median tells you what a group of homes sold for; it does not tell you what yours is worth, because it knows nothing about your floor, your outlook, your renovation or your lease. They are not advice, and they are not a forecast — every figure here is a record of what has already happened.

If you want a figure for your own home rather than for its estate, start here.