Private Home Prices Q2 2026: The Index Rose 0.5%. The Median Fell 6%. Both Are Right.

In brief: URA's private residential price index rose 0.5% in Q2 2026, its slowest gain in three quarters. Yet the median transacted price of a non-landed home fell over the same period. Both are true, and the gap between them is the story. The index holds quality constant; the median does not, so when the mix of what sold shifts toward cheaper suburban resale and away from prime new launches, the median drops even as like-for-like prices edge up. Underneath the headline: landed homes surged while condos stalled, the prime core outperformed the suburbs for once, and resale took 62% of the market. If you are buying, selling or upgrading, the segment you are in matters far more than the index.

---

What URA reported

The private residential property index rose 0.5% in Q2 2026, cooling from 0.9% in Q1. The headline masks a sharp split by property type: landed homes rose 2.5% (reversing a 0.4% fall the quarter before) while non-landed homes slipped 0.1% (down from 1.3% growth). The market did not move as one.

Within non-landed, the regions inverted the recent pattern. The Core Central Region (CCR) rose 1.8% — prime central prices strengthening — while the Rest of Central Region (RCR) fell 1.2% and the Outside Central Region (OCR) eased 0.1%. For once the expensive core outperformed the city fringe and the suburbs, the same shape playing out in the HDB market, where central and mature towns pulled ahead of the heartlands (our HDB Q2 brief).

Transaction activity rose. Developers sold 2,141 new private homes (excluding executive condominiums), up from 2,013 in Q1. Resale volume jumped to 3,813 from 3,225, and resale made up 62% of all sale transactions — buyers continuing to favour completed homes over new launches. Sub-sales added 194.

Source: URA.

Why the median tells a different story

The index above is quality-controlled: URA measures price change for a constant basket, so it isolates genuine price movement. A raw median does not. It simply reports the middle transaction, so it swings with whatever happened to sell that quarter. In Q2 those two lenses point opposite ways, and the reason is instructive.

By Realila Research's count of caveats lodged to date, the median non-landed transacted price eased quarter-on-quarter — even though URA's index says non-landed prices barely moved. The gap is mix. New launches, which cluster in expensive prime districts, made up a far smaller share of Q2's transactions than Q1's; resale and suburban homes made up more. Cheaper homes formed a larger slice of the pie, so the median slid while like-for-like values held. This is not the market repricing down. It is the market's composition shifting, and it is exactly why the index, not the median, is the right tool for "did prices move."

We show the levels below because they tell you what people actually paid and where the volume went. We do not present them as price movements — the index already answers that, and reading a mix-driven median as a price change is the single most common error in quarterly property commentary.

Where the volume actually went

Non-landed transaction levels by region, Q1 to Q2 2026, caveats lodged to date. Read these as activity and level, not price movement — the constant-quality direction is URA's index above.

Region	Q1 txns	Q2 txns	Q2 median PSF
CCR (prime core)	1,199	753	S$2,348
RCR (city fringe)	1,290	1,639	S$2,178
OCR (suburbs)	2,167	3,013	S$2,025

Source: URA & Realila Research. Levels reflect caveats lodged to date and the mix of what sold; the quality-controlled price direction is URA's regional index (CCR +1.8%, RCR −1.2%, OCR −0.1%).

The shift is stark: CCR's share of activity shrank while OCR's swelled. Fewer prime launches, more suburban resale — precisely the mix change that drags a raw median down while URA's index shows CCR prices actually firming. The suburbs did the volume; the core did the price.

Landed broke away

The clearest signal in the quarter is the divergence between landed and non-landed. Landed homes — houses, not condos — rose 2.5% on URA's index, swinging from a decline, while non-landed essentially flatlined. Landed supply is fixed and shrinking as a share of the market, and demand for it is far less elastic than for condos, where a heavy pipeline of new launches gives buyers options. When the overall "private market up 0.5%" line is really "houses surged, condos stalled," the averaged headline hides more than it tells.

For upgraders eyeing a landed home, the message is that this segment is not tracking the broader cooling; for condo buyers, the flat non-landed index means negotiating room persists, especially in the RCR where the index actually fell.

The EC drought

Executive condominiums — the hybrid public-private format upgraders lean on — barely featured in Q2: no new EC projects launched, and only 175 units sold across the quarter, down sharply from Q1's launch-driven surge. With the pipeline thin, EC scarcity is a live constraint for the exact buyers who most depend on the format to bridge from HDB to private. It is a supply story worth watching into the October BTO exercise and the next EC tenders.

The supply wave building

URA's supply figures point to a market loosening ahead, not tightening. Around 9,320 units sit on the Confirmed List for 2026 — more than 50% above the ten-year average — and roughly 60,600 units are due to complete over the coming years. The vacancy rate for completed private homes ticked up to 6.4%. Rents rose 0.7%, cooling from the prior quarter. A market absorbing record supply while prices merely level is a market where buyers gain leverage over time, not lose it.

What this means for your decision

If you are selling. Which lens applies to your home? A landed seller is in a strengthening segment — price to the recent move up. A condo seller is in a flat market where the index barely rose and the RCR fell, so price to current transactions, not to last year's peak, or expect to sit. And do not be spooked by "median prices fell" headlines: that is the mix, not your home's value.

If you are buying. Negotiating room is widest in non-landed RCR stock and where the incoming supply is heaviest. The core is firming, so prime buyers have less leverage than the "market cooling" narrative suggests. Match your expectation to your segment, not to the national headline.

If you are upgrading. Landed strength and the EC drought both cut against you if that is your target; the flat condo market cuts for you. Either way the binding number is usually stamp duty. Before you commit to a buy-first or sell-first sequence, compute your exact BSD and ABSD position with the Realila stamp duty calculator. It is free and takes a minute.

What to watch next

Whether the CCR's outperformance holds — one quarter of prime strength is a data point, not a trend. Whether landed keeps breaking away from non-landed, which would mark a structural split rather than a blip. And how the market absorbs the supply wave: 9,320 Confirmed List units and rising vacancy into a quarter where prices already only levelled. The Q3 flash estimate, out in early October, is the next read.

Coming to Realila Research: freehold vs leasehold, on live data. The price gap, how it has moved, and whether the leasehold discount is widening — built from the transaction record, refreshing as new caveats land. Get notified when it opens.

---

This brief is general market research, not financial advice, and figures are dated to the periods they cover.

Data and methodology

Price direction, index changes, official transaction counts, and supply figures are URA's, from the Q2 2026 quarterly statistics. Transaction levels, regional volumes and median PSF are Realila Research's, computed on URA REALIS caveat data for private residential transactions with contract dates in the stated quarters. Q2 2026 figures reflect caveats lodged to date; because caveats lodge over the weeks and months following a transaction, the most recent quarter is provisional and its volumes will rise as more caveats are recorded — our Q2 resale count sits below URA's published total for this reason. Medians are of raw transacted prices, not quality-controlled, so they reflect both price movement and the mix of what sold within each cut; we present them as levels and activity, not as price changes, for which URA's index is the correct measure. Executive condominiums are excluded from non-landed figures to match URA's headline convention except where stated. Regional classification (CCR/RCR/OCR) follows the standard URA district approximation.