By Winfred Quek · CEA R073319H · Published 28 October 2026
Facts verified: 28 October 2026 · Sources listed below
Most sellers and buyers search "how much is my HDB worth" and land on a portal estimate or a friend's number. HDB itself publishes the raw material every estimate is built from: a searchable record of registered resale prices, a downloadable dataset, and one quarterly index. This page shows how to read those three sources, step by step, using a worked example I computed from the open dataset. It is a reading guide, not a valuation. For how asking prices are set once you have comparables, see how to price your HDB flat to sell, and for the official valuation that comes later in the sale, see the HDB Request for Value fee and steps.
The three official sources, and what each is for
- Check Resale Flat Prices (HDB online service). Good for a quick look at recent registered sales in one town or street. HDB describes it as a way to "search and view recent transacted prices of resale flats".
- The data.gov.sg resale dataset. The same type of records as a file you can filter yourself, with more columns, including storey range and remaining lease.
- The HDB Resale Price Index (RPI). One number per quarter for the whole resale market. It is context, not a price for any flat.
Source: HDB, Check Resale Flat Prices, checked 28 October 2026. https://services2.hdb.gov.sg/webapp/BB33RTIS/
Step 1: use Check Resale Flat Prices for a first look
On the day I checked, the search form asked for a flat type (required), then either an HDB town or a street name, with an optional block number range. It also asks for a resale registration date range, with choices of the last 6 months, the last 12 months, or a custom range of up to one year, and an optional range for the transacted price. You can also state whether you are enquiring as a seller or a buyer.
HDB's Flat Portal links to the same service from its resale listings, noting that it covers "transactions over the past 2 years". The listing pages also show a price range of nearby flats of the same type sold within 200 metres in the past 6 months. Those are useful pointers, but a range is not a median, and it does not tell you how many sales sit behind it.
Source: HDB Flat Portal, resale listing details page (example listing, may be removed), checked 28 October 2026. https://homes.hdb.gov.sg/home/resale/42374
What this service does not let you filter by is storey range or remaining lease. For those, use the dataset.
Step 2: download the dataset and filter it yourself
The dataset is called "Resale flat prices based on registration date from Jan 2017 onwards", managed by HDB and published on data.gov.sg. When I checked, it covered January 2017 to October 2026 and had been updated that same day. Each row has the month of registration, town, flat type, block, street name, storey range, floor area in square metres, flat model, lease commencement date, remaining lease, and resale price.
HDB's notes on the dataset matter for how you read it. The floor area includes any recess area purchased and similar additions. The transactions exclude resales that may not reflect the full market price, such as resale between relatives and resale of part shares. And HDB says resale prices "should be taken as indicative only", because the price agreed between a buyer and seller depends on many factors.
Source: HDB via data.gov.sg, Resale flat prices based on registration date from Jan 2017 onwards, dataset notes, checked 28 October 2026. https://data.gov.sg/datasets/d_8b84c4ee58e3cfc0ece0d773c8ca6abc/view
A practical way to use it:
- Download the file and open it in a spreadsheet.
- Filter town and flat_type to match your flat.
- Filter month to a window you can state, for example the last 12 full months. A stated window keeps you honest about how old your comparables are.
- Narrow by storey_range and remaining_lease, watching how many sales are left. When a filter leaves fewer than about 30 rows, treat the median as a rough guide only.
- Take the median, not the average, and look at the spread between the 25th and 75th percentile as well.
Notice what is missing from the columns. There is no field for renovation condition, facing, view, the exact floor within the storey band, or the HDB valuation. Those are the things that explain why two similar flats sell apart.
Worked example: 4 room flats in Tampines
To show the steps, I filtered the dataset for Tampines, 4 ROOM, with registration months from October 2025 to September 2026 (the latest 12 full months in the file when I checked). This is a labelled illustration of method. It is not a valuation of any flat, and a different town, flat type or window will give different figures.
| Filter step | Sales left | Median price | Middle half of sales |
|---|---|---|---|
| 1. Tampines, 4 ROOM, Oct 2025 to Sep 2026 | 902 | $665,000 | $620,000 to $760,000 |
| 2. Add storey range 07 TO 09 or 10 TO 12 | 368 | $668,000 | $633,000 to $770,000 |
| 3. Also remaining lease from 60 years up to 75 years | 141 | $650,000 | $626,400 to $675,000 |
Source: author's calculation from the HDB resale dataset on data.gov.sg, rows registered October 2025 to September 2026, retrieved 28 October 2026. "Middle half" means the 25th to 75th percentile.
Two things stand out. First, filtering to a similar storey band changed the median very little, but narrowing the lease band tightened the spread a great deal: the middle half of sales went from a $140,000 span in step 1 to under $50,000 in step 3. Second, remaining lease alone moved the picture sharply. Across all floors, the same 12 months gave these medians:
| Remaining lease at sale | Sales | Median price |
|---|---|---|
| Under 60 years | 234 | $619,444 |
| 60 years to under 75 years | 320 | $640,000 |
| 75 years and above | 348 | $775,000 |
Source: author's calculation, same dataset and window as above.
That is why a town level median can mislead. A single "Tampines 4 room" number blends flats with very different leases. If you compare a flat with a much shorter or longer lease to the town median, you are comparing different products. Storey and lease can also overlap in the data, so check which sales sit behind a high floor result before you read it as a floor premium.
What the Resale Price Index is, and is not
HDB says the RPI "tracks the overall price movement of the public housing market". The series on data.gov.sg uses 1Q2009 as 100. The Department of Statistics notes that the index is based on the quarterly average resale price by date of registration, and that from 4Q2014 it has been computed with the stratified hedonic regression method.
The latest quarters in the dataset, when I checked on 28 October 2026, were:
| Quarter | RPI (1Q2009 = 100) |
|---|---|
| 3Q2025 | 203.7 |
| 4Q2025 | 203.6 |
| 1Q2026 | 203.4 |
| 2Q2026 | 202.8 |
| 3Q2026 | 202.5 |
Source: HDB via data.gov.sg, HDB Resale Price Index (1Q2009 = 100), Quarterly, dataset last updated 27 October 2026. https://data.gov.sg/datasets/d_14f63e595975691e7c24a27ae4c07c79/view
Read it carefully:
- An index level of 202.5 is not a price in dollars. It says the market is at about 2.0 times its 1Q2009 level.
- It covers all towns and flat types together. It cannot tell you what a 4 room flat in one town did.
- It is quarterly, so it lags. Over the five quarters above, the index stayed within about 1.2 points of itself, which says the overall level has been broadly flat, not that your flat's value has not changed.
- It is not a forecast. This page does not predict where prices go next.
Sources: HDB, Resale Statistics, https://www.hdb.gov.sg/managing-my-home/selling-a-flat/process-for-selling-a-flat/resale-flat-planning/resale-statistics ; Department of Statistics, HDB Resale Price Index table notes, https://tablebuilder.singstat.gov.sg/table/TS/M212161 . Both checked 28 October 2026.
What these sources cannot tell you
Open data gives you a defensible range of what similar flats registered at. It does not give you the price your own flat will fetch, and it does not replace the valuation. When a buyer uses CPF or a housing loan, HDB sets a valuation through the Request for Value. The valuation process is explained in our Request for Value guide. If you want to see how a seller can test that figure, see getting a second opinion before listing.
For town level starting points built the same way, our pages on what $600,000 buys by town and the cheapest towns for a 4 room flat use the same dataset, and the Tampines price page shows how a single town looks by flat type.
A short checklist before you rely on any number
- State the window (for example Oct 2025 to Sep 2026) and the count of sales behind your figure.
- Match flat type, town, storey band and remaining lease, in that order.
- Use the median and the middle half, not the single highest or lowest sale.
- Remember the data has no condition, facing or valuation column.
- Use the RPI only as market context, never as a flat price.
Frequently asked questions
Where can I see what HDB flats in my area actually sold for?
Use HDB's Check Resale Flat Prices online service for recent registered sales by flat type, town or street, and a date range of up to one year at a time. For storey range and remaining lease as well, filter the resale flat prices dataset on data.gov.sg, which is managed by HDB. Both show registered resale prices and are indicative only.
Is the HDB Resale Price Index the price of my flat?
No. The Resale Price Index tracks the overall price movement of the public housing market, with 1Q2009 set to 100. It is a quarterly market wide number, not a dollar price, and it cannot tell you what one flat in one town is worth.
Why do two flats of the same type in the same town have different prices?
Storey, remaining lease, condition, facing and block location all differ, and the dataset records only some of them. In the Tampines 4 room example, median prices ranged from $619,444 for leases under 60 years to $775,000 for leases of 75 years and above, over the same 12 months.
Can I use these figures as my HDB valuation?
No. They are registered sale prices for other flats. The HDB valuation is a separate step in the resale process, and the price a buyer and seller agree on depends on many factors. Use the data as a reference range and take advice on your own flat.
Want recent sales for your own flat type and town read with you?
Share your town, flat type, storey and lease, and Winfred will go through the recent registered sales with you so you know what range the data shows. General information only, not a valuation.
Ask Winfred on WhatsApp Book a free 30 minute callWinfred Quek, CEA R073319H, Crestbrick Pte Ltd (L3010886H). General information as at 28 October 2026, not legal, financial, tax or property advice. Figures are from HDB, data.gov.sg and the Department of Statistics, and may change. The worked example is the author's calculation from the public dataset. This page makes no price forecast.