Reading the data
How to read HDB resale prices: median, average and price per square foot
By HDBData · Figures are recalculated every week from HDB’s records on data.gov.sg · How the numbers are worked out
A headline such as “HDB resale prices hit a new high” can be describing several different numbers. It might be a median, an average, a price index or a price per square foot, and each one answers a different question. This guide explains the figures HDBData shows, using the 24,873 resale transactions recorded in the 12 months to November 2026.
Median and average are not the same number
Line up every flat sold in the last 12 months from cheapest to dearest. The median is the price of the one in the middle: $635,000. Half the flats sold for less and half for more. The average adds up every price and divides by the number of sales, which gives $666,592.
The average is $31,592 higher, about 5%. That gap is not an error. A flat can sell for $1.5 million but no flat sells for less than nothing, so the expensive end of the market stretches further from the middle than the cheap end does, and a small number of very expensive flats pull the average up. The median ignores how far out those sales are. It only cares which side of the middle each sale falls on.
This is why HDBData reports medians throughout. On a single street, where a year might bring only 20 sales, one unusually large maisonette can move an average by tens of thousands of dollars without the typical flat changing at all.
A median still hides the spread. The middle half of all sales fell between $515,000 and $788,000, so a quarter of buyers paid less than the lower figure and a quarter paid more than the higher one. A flat priced well away from the median is not necessarily mispriced; it may simply be a different kind of flat.
A median depends on which flats happened to sell
The overall median mixes every flat type together, and the types are priced very differently.
| Flat type | Share | Median price | Psf |
|---|---|---|---|
| 2-roomtypically 47 sqm | 3% | $375,000 | $762 |
| 3-roomtypically 67 sqm | 22.8% | $440,000 | $590 |
| 4-roomtypically 93 sqm | 43.4% | $630,000 | $615 |
| 5-roomtypically 116 sqm | 24.2% | $743,000 | $581 |
| Executivetypically 146 sqm | 6.6% | $915,000 | $586 |
Psf is the median price per square foot. 1-room and multi-generation flats are left out: only a handful sell each year.
4-room flats made up 43.4% of sales, more than any other type, so the overall median of $635,000 sits close to the 4-room median of $630,000. If the mix shifts, the overall median shifts with it even when no individual flat has become dearer. A quarter in which more 5-room and executive flats are sold will show a higher median than a quarter of mostly 3-room sales.
The same effect distorts comparisons between towns. Take Pasir Ris: its overall median is $730,000, well above the national $635,000. But its 4-room flats sold for a median of $635,000, against $630,000 for 4-room flats across Singapore. Pasir Ris looks expensive mainly because 58.6% of the flats sold there were 5-room or executive flats, compared with 30.7% nationally. The guide to the cheapest and most expensive towns ranks every town both ways.
The practical rule: compare a flat with others of the same flat type. On HDBData, every town and year page breaks the median down by flat type for this reason.
What price per square foot is for
Price per square foot (psf) divides the resale price by the flat’s floor area. HDB records floor area in square metres, and one square metre is about 10.764 square feet, so a 93 sqm flat is about 1,001 sq ft. Psf lets you compare two flats of slightly different sizes: a 4-room flat of 90 sqm against one of 100 sqm.
It works less well across flat types. 2-room flats sold for a median of $762 per sq ft, while 5-room flats sold for $581. Size is part of the reason: every flat needs a kitchen, bathrooms and a front door however small it is, and more buyers can afford a lower total price. Age is the other part. The typical 2-room flat sold had a lease that began in 2018, the typical 5-room flat in 2000. A psf gap between two flat types is therefore a mix of size, age and location, and says little about which is better value.
Psf is most useful when the flat type is held fixed. Within one town, the psf of 4-room flats on different streets, different floors or different lease ages tells you what buyers actually pay extra for. The guides on floor level and remaining lease both use it that way.
The newest months are always incomplete
HDB adds a resale to the public dataset when the transaction is registered, and keeps adding sales to a month for weeks after that month has ended. The newest month in the data is November 2026, with 1,027 sales so far. A typical recent month ends up with about 2,085, so it is roughly 49% complete.
Two things follow. A sharp drop in the number of sales in the latest month is almost always the data filling in, not the market slowing. And a median taken from a part-filled month can swing by a few percent once the rest of the sales arrive. HDBData therefore uses 12-month windows for its headline medians, and treats the current year as incomplete on the year pages.
What a resale record does not tell you
Each record gives the month, town, block, street, flat type, flat model, floor area, a storey range, the year the lease began and the price. That is a lot, but several things that move a price are missing:
- The exact unit. Storey is given as a range of three floors, such as 10 to 12, and there is no unit number. Two sales in the same block and range may face a park and a car park.
- Condition. A newly renovated flat and one untouched for 30 years appear identically.
- How the price was made up. The record shows the final resale price only. It does not show the valuation or how much of the price was paid in cash above it.
- Asking prices. These are completed transactions. What sellers are asking today can be higher or lower.
Checking a price in practice
To judge whether a price for a particular flat is reasonable, narrow the comparison step by step rather than relying on a town median:
- Start with the search: enter the block and street, and pick the flat type.
- Look at sales in the same block in the last 12 to 24 months, then widen to the street if there are fewer than five.
- Check the storey range and floor area of each sale against the flat you are looking at, and compare psf rather than total price when sizes differ.
- Note the lease start year. Neighbouring blocks can be 20 years apart in age.
The methodology page sets out exactly how every figure on the site is calculated.