The $45 million median was one street, not the suburb.
My top ten most expensive Sydney suburbs put Point Piper at a median house price of $45 million. That figure is real and it belongs to Wolseley Road, the single most expensive street in the country. The same source that published it gave Point Piper’s suburb median as $15.2 million on the same page. I took the street number and attached it to the suburb, which is a forty-five million dollar mistake at the top of a list that then ranked nine other suburbs underneath it. The list also included a suburb that is not in the Eastern Suburbs, and concluded that these markets “continue to outperform” while reporting one of them down 17.4 per cent.
The headline figure, the geography, and a conclusion the list itself disproves.
This was a round-up of other people’s data, which is exactly the kind of article where errors travel furthest, because a reader has no reason to doubt a table.
A street median, not a suburb median
Forty-five million is the published median for Wolseley Road, described in the same source as the nation’s priciest street. That source gave Point Piper’s suburb median as $15.2 million. Cotality’s Best of the Best for 2025 put it at $17 million on four sales; its September 2026 figure is $13.47 million. The $45 million sat at the top of a ranked list and set the reader’s expectation for everything below it.
Correctly described, wrongly included
The entry itself says north shore. Woolwich sits on the Hunters Hill peninsula, about nine kilometres from the city on the other side of the harbour, and it has since dropped out of Sydney’s top ten entirely: on Cotality’s September 2026 ranking the only non-Eastern Suburbs entrant is Cremorne Point. Including it was not wrong as a fact about Sydney. It was wrong on a page written for Eastern Suburbs readers without a word of explanation.
Contradicted by the list above it
Double Bay appears at minus 17.4 per cent over twelve months and Dover Heights units at minus 53.1 per cent. Two of the ten entries were falling, one by more than half, and the closing paragraph says they are outperforming. A summary should describe the table, not override it.
Three providers, no attribution
Those three measure different things on different samples over different windows. In a category where one suburb’s published median ranges from $13.5 million to $55 million depending on who you ask, listing three sources at the top and then giving a single figure per suburb is not a citation. It is the appearance of one.
Accurate and meaningless
Centennial Park has almost no apartment stock, so the figure moves on a handful of sales and sits below the Sydney unit median. Current published figures for the same suburb run from about $870,000 to $1.01 million. Printing a sub-million apartment median inside a most-expensive-suburbs list is the kind of detail that tells a careful reader the table has not been thought about.
Four sources, one suburb, $2.86m apart: the recovery call scored.
Every verified sale on every street: Bellevue Hill by street.
Every figure below is a published 2026 median for the same suburb.
None of these providers is wrong. They use different windows, different registers and different rules about what counts, and in a suburb with a handful of sales those choices decide the answer.
One consistent source, one date. It is the only honest way to rank anything.
Below is the top ten on Cotality’s September 2026 median house values, which is a single provider on a single date, with Bellevue Hill and Vaucluse taken from mid-2026 readings of the same source because they were not in that release. The dashed bars are the suburbs that fell since November 2025. Note where Point Piper lands, and note that Rose Bay is the only one of the eight comparable suburbs that is higher than a year ago.
Nine of the ten are in the Eastern Suburbs, which is the genuinely remarkable fact and the one the original list buried under a wrong number. The other thing worth sitting with: eight of these suburbs appeared in the November 2025 ranking too, and seven of the eight are lower now. The suburb that was supposed to be most insulated, Point Piper, has fallen the furthest, and by a margin.
Point Piper has 218 properties. Four of them sold.
The more expensive the suburb, the fewer the sales, and the less a median means. This is not a flaw in anyone’s data. It is arithmetic, and it is strongest exactly where the headlines are.
One specific house
With seven sales, the median is the fourth one, ranked by price. Not an average, not a model: one house that happened to change hands. Everything said about the suburb on the back of that figure is really a statement about that house and the three either side of it.
The biggest deals are often missing
At this level many transactions settle off-market with confidentiality agreements and never reach the register at a stated price. Cotality has described exactly this problem in Point Piper. So the published median is a median of the sales that were disclosed, which at the top end is a self-selecting sample.
Comparable sales, with dates
For any individual property the only numbers that mean anything are three or four genuinely comparable sales on streets you would accept as equivalent, with the condition and land size allowed for, and a date on each one. A suburb median is background. It has never valued anybody’s house.
Set the number of sales and the spread. The tool tells you what the median is standing in for.
Enter how many houses sold in your suburb over the last year and the range those sales covered. It works out which single sale sets the median, how much of the suburb sits outside the middle, and whether the figure is worth quoting at all.
The reliability bands are my own rule of thumb from thirty-five years of campaigns, not a statistical test: under ten sales an anecdote, ten to thirty indicative, thirty to eighty usable, above eighty reasonably stable. Even a stable median says nothing about an individual property’s condition, land or street.
That is the honest summary. Forty-five million dollars sat at the top of a ranked list on my own site, and it was Wolseley Road, not Point Piper. The number was not invented and it was not even wrong in its original context. It was simply attached to the wrong thing, which is how most bad property data gets into circulation: a correct figure, moved one level up or down, and then repeated by everyone who reads it.
What I have changed my mind about is publishing suburb rankings at all. The mathematics at the top of this market defeats the format. Point Piper has 218 properties and four recorded sales in a year, which means its median is one house, and the published figures for it this year run from $13.5 million to $55 million. You cannot rank things you cannot measure, and dressing four transactions up as a suburb median gives a reader a precision that does not exist. If I run this list again it will carry sale counts beside every figure and one named source with one date, or it will not run.
The genuinely interesting fact got lost under all of this. Nine of Sydney’s ten most expensive suburbs are within a few kilometres of each other in the Eastern Suburbs, and seven of the eight comparable entries are worth less than they were a year ago. That is a real story about concentration and about a correction reaching the very top of the market for the first time in a decade. It did not need a forty-five million dollar headline to be worth reading.
Trophy sales checked against the register: four sales, four wrong dates.
Scorecards, suburb guides and street data: Alan’s articles.
Your house is not a median. Let me show you the sales that actually bear on it.
Three or four genuinely comparable sales on streets you would accept as equivalent, with dates, condition and land size allowed for, and an honest view of what a campaign would do right now. That is the only number that has ever valued a property. Thirty-five years in the Eastern Suburbs, every campaign run personally.
35+ years in Sydney’s Eastern Suburbs
- Rose & Jones, Most Expensive Sydney Suburbs 2025: Wolseley Road in Point Piper has a median house price of $45 million, making it the nation’s priciest street; the same guide gives Point Piper’s suburb median house price as $15.2 million.
- Domain, reporting Cotality’s Best of the Best 2025: Point Piper’s median house price $17 million, with four sales recorded in the twelve months to September against 72 in Bellevue Hill, whose median was $11,604,624; Point Piper has 218 properties and a turnover rate of 2.3% against a Sydney average of 4.3% and a national average of 4.8%. Cotality’s Tim Lawless notes Point Piper frequently does not meet the filter criteria for inclusion because so few sales occur, and that confidential transactions mask the depth of the market.
- Cotality median house and unit values, September 2026, as compiled in a published top-ten ranking: Point Piper $13,471,273 and $2,824,971; Darling Point $9,146,863 and $2,571,044; Centennial Park $7,407,813 and $911,944; Rose Bay $6,976,964 and $1,540,018; Double Bay $6,627,520 and $1,743,262; Dover Heights $6,378,594 and $1,290,835; Tamarama $6,198,009 and $1,763,112. Of the eight suburbs common to this list and Cotality’s November 2025 top ten, Rose Bay was the only one higher, up about 4.4%; Dover Heights and Tamarama fell least at about 2.5% and 3.7%; Point Piper fell 22.2%. Cremorne Point is the only Lower North Shore entrant.
- Duotax, using Carto Australia median sale prices for the twelve months to September 2026: Point Piper $25,000,000 house and $4,600,000 unit, the house figure based on eight sales; Darling Point $9,130,000 and $2,650,000; Double Bay $8,200,000 and $2,300,000; Tamarama $8,070,000 and $1,880,000; Dover Heights $7,400,000 and $1,530,000; Rose Bay $6,880,000 and $1,850,000.
- The Ben Collier Team, Eastern Suburbs market report, July 2026, using Cotality figures to May 2026: Double Bay median house price $8,400,000, up 21.7% over twelve months; Rose Bay $5,925,000, up 11.8%; Centennial Park $13,640,000, up 117.4% over twelve months; Centennial Park median apartment price $870,500.
- Rose & Jones Eastern Suburbs Property Guide, 2026: Double Bay median house $7,850,000, up 27.2%, units $1,825,000; Dover Heights median house $6,850,000, up 11%, units $1,450,000, up 74.7%; Rose Bay median house $6,000,000, units $1,574,250; Centennial Park median house $4,412,500, units $1,010,000.
- Further published 2026 figures for Point Piper: realestate.com.au $25 million for February 2025 to February 2026; a Newcastle broker comparison at $26.5 million for June 2026; realsearch.com.au at $55 million, described as highly variable due to low volume; auspropertyinsights at $31.5 million. Weiss Real Estate suburb page, April 2025 to March 2026: $39.5 million as the median of three recorded house sales.
- Everstone Finance, June 2026, citing Cotality: Rose Bay house median about $5,270,000 and unit median about $1,930,000. Levy Property Group, 2026: Rose Bay house median around $5.75 to $5.9 million.
- Weiss Real Estate, Bellevue Hill Property Prices by Street 2020–2026 (verified sales register): house median $12.59 million in 2025 falling to $9.78 million across the first seven months of 2026. propertyvalue.com.au, twelve months to late 2026: Vaucluse median value $9.4 million on 89 sales with an average vendor discount of 10.3% and the median down 5.0% over the year.
- Alan Weiss, “Sydney’s 10 most expensive suburbs” (the original article this corrects), and Weiss Real Estate campaign records, Eastern Suburbs.
Every figure on this page names its source and its date because the providers disagree substantially. Median sale prices, hedonic median values and asking-price medians measure different things and are not interchangeable. In suburbs with few annual sales, a median reflects which properties happened to transact rather than what properties are worth, and should not be used to value or price any individual property. Calculator outputs are illustrative arithmetic on figures you enter. Includes Alan Weiss’s personal views. General information only; not financial, tax or legal advice.