HomeResearch › MRT premium
Updated 11 Jul 2026 · 70,665 resale caveats (projects ≥5 yrs) · OneMap/SLA

How much is walking distance to an MRT really worth?

TL;DR
+13%
≤300 m from MRT
+7%
300–500 m
~0%
beyond ~500 m (fades out)
70k
caveats analysed

The walkability curve

“Near the MRT” is the most-used line in a sales gallery — but the market doesn’t treat 200 m and 900 m the same. Holding local neighbourhood (~2km cluster), tenure, unit size and sale timing constant, here is the $psf premium buyers actually pay by distance to the nearest station, versus a project more than a kilometre away:

$psf premium vs a project >1 km from MRT

Hedonic estimate, all private caveats. Bars are the premium each distance band commands over the >1 km baseline.
+0%+5%+10%+15%+13%≤300m+7%300–500m+0%500–800m
Key finding: the premium is front-loaded. It concentrates in the first ~300 m (+13%), roughly halves by 500 m, and is essentially gone beyond ~500 m. Walkability is real value; vague “proximity” is not.

Where MRT access matters most

The premium isn’t uniform. It is steepest in the city-fringe RCR (+12% within 300 m), where an MRT-adjacent home is a prime, highly rentable location; it is weakest in the prime CCR (+9%), where the most exclusive addresses sit deliberately away from the bustle of a station:

Segment≤300 m300–500 m500–800 m
CCR+9%+7%+4%
RCR+12%+8%+-1%
OCR+13%+7%+-1%

What the data means

The market pays for true walkability, not for a postcode that happens to list a station name. The strongest, most defensible premium sits inside ~300–500 m — an easy walk in the heat — and it is worth most where the station links to jobs or an interchange. Beyond ~800 m, “near MRT” is marketing, not value: the resale market will not reward it, so neither should your entry price.

Investor verdict

Pay up for MRT proximity only when it’s a genuine ≤500 m walk, ideally to an interchange or a job-node line — that is where the market reliably rewards it (up to ~+13%). Do not pay an MRT premium for a project 800 m+ away, however the brochure frames it; the buyer who follows you won’t.

Future-line upside is a separate bet — the question is how much is already priced in before the line opens. Check the URA Master Plan for the alignment and confirm the actual walking route (not the straight line) before you decide.

How to use this before you buy

Measure the real walk, not the straight line

Our distances are straight-line; the walking route can be longer (across a canal, a main road, no sheltered link). Check the actual path — a “400 m” project can be a 700 m walk.

Match the premium to the band

If you’re paying a visible MRT premium, confirm the project is genuinely inside ~500 m. Beyond that, discount the pitch and price it like any other project in the district — check comparables in the Valuation tool.

Weigh the line, not just the station

Proximity to an interchange or a job-node line (to the CBD, JLD, one-north) is worth more than a quiet branch stop. See where jobs are heading on growth areas.

For a future line, ask what’s priced in

If the upside is a station opening in a few years, check how much the market has already paid ahead of it — and read the district page for current medians and trend.

How is this worked out? — hedonic model, controls, sources & limits
Sample
70,665 resale caveats in projects completed at least 5 years ago (by 2021) — so the premium reflects the settled resale market, not developer launch pricing. Each matched to its mapped nearest-MRT distance.
Model
A hedonic OLS of log($psf) on distance-band dummies (≤300 / 300–500 / 500–800 / 800–1000 m, baseline >1000 m), with local ~2 km cluster fixed effects (finer than district, so within-neighbourhood distance to the station identifies the premium), a freehold dummy, log(unit size) and a time control — so the band premium isn’t confounded by location, tenure, size or when the unit sold.
Scope
Market-level (all districts) plus a CCR/RCR/OCR split. Results are averages; a specific project can differ.
Sources
URA Data Service caveats (past ~5 years); nearest-MRT distances from station coordinates via OneMap/SLA. $psf outliers & en-bloc rows excluded.
Limits: straight-line distance to the nearest existing station (not walking route, and future lines are not separated — a follow-up study). The premium is an average conditional on the controls, not a valuation of any unit.
For educational purposes only — not financial advice.
Keep going: all research › · value a specific unit › · growth areas & future lines ›