Transaction vs asking prices — what apartments in Warsaw actually sell for

In June 2026, the median asking price for an apartment in Warsaw is 17,658 zł/m². The median transaction price — the one written into notarial deeds and filed in the state registry — is 15,323 zł/m². The gap: 15.2%.

Does that mean you can knock 15% off any Otodom listing? No. And that's what this article is about — because the gap between asking and transaction prices is one of the most quoted and most misunderstood numbers in real estate. We'll show real data for all 18 Warsaw districts, explain where the gap comes from, where the statistics lie (watch out for Żoliborz), and how to turn these numbers into negotiating leverage — instead of waving a median around like a sabre.

One city, two price lists

Poland's housing market runs on two parallel sets of prices.

Asking prices are the ones everyone sees: Otodom, OLX, developer websites. They're public, fresh and… aspirational. They're what the seller would like to get — sometimes realistic, sometimes padded "to leave room for negotiation", sometimes detached from reality because "the neighbour sold for that much in 2024".

Transaction prices are hidden in notarial deeds. Every property sale in Poland is filed in the RCN — Rejestr Cen Nieruchomości (Property Price Registry), maintained by county offices and cities with county rights. These are the amounts actually paid, from documents bearing a notary's seal. The catch: the registry isn't served on a platter. Data arrives with a lag of several months, and access requires a formal request to the county office or paid tools.

This creates the information asymmetry that buyers regularly pay for: the seller sets the price by looking at other listings (i.e. other people's wishes), while the buyer has no idea that three similar apartments in the same building just sold for 8% less.

The real numbers: 18 Warsaw districts, June 2026

The table below compares the median transaction price (RCN, apartments, last 12 months of registered deeds — about 13,700 transactions) with the median asking price of active listings (about 7,600 listings, as of June 2026).

District	Median transaction (zł/m²)	Median asking (zł/m²)	Gap
Bemowo	13,902	16,700	+20.1%
Wilanów	17,922	20,401	+13.8%
Włochy	15,780	17,600	+11.5%
Wesoła	12,911	14,341	+11.1%
Ursus	13,422	14,820	+10.4%
Ochota	16,421	17,970	+9.4%
Mokotów	17,617	19,150	+8.7%
Praga-Południe	16,260	17,573	+8.1%
Wawer	12,907	13,926	+7.9%
Wola	19,036	20,521	+7.8%
Białołęka	13,182	14,104	+7.0%
Śródmieście	20,507	21,480	+4.7%
Praga-Północ	17,827	18,513	+3.8%
Targówek	14,211	14,608	+2.8%
Rembertów	12,931	13,252	+2.5%
Bielany	16,030	16,318	+1.8%
Ursynów	16,826	16,904	+0.5%
Żoliborz	23,606	20,370	−13.7%

Three things jump out.

First: the spread is enormous. From +20% in Bemowo to +0.5% in Ursynów. A city-wide "average asking-to-transaction gap" is a nearly useless number when you're looking at a specific apartment — a typical district sits around +7-8%, but the extremes differ from that several times over.

Second: Żoliborz has a negative gap. The median asking price is lower than the median transaction price by almost 14%. Are Żoliborz sellers paying buyers to take their apartments? Obviously not — we'll unpack what this number actually says in a moment, because it's the most important methodological lesson in this article.

Third: these are medians, not pairs of "the same apartment in the listing and in the deed". We're comparing the structure of supply with the structure of sales — and they can differ. That's why part of the gap isn't negotiating margin at all, just statistical apples and oranges.

Where does the gap come from? Four mechanisms

1. Genuine negotiating margin

The simplest component: sellers list higher than they're willing to accept. They leave themselves a 3-7% buffer because "everyone haggles in Poland". Buyers sense it, bid below asking, and the deal closes somewhere in the middle. This mechanism is real — but it explains only part of the observed gap.

2. What's listed ≠ what sells

The asking median describes what's sitting on the market; the transaction median describes what sold. Different baskets.

The best example: Bemowo (+20.1%). Can you negotiate a fifth off the price in Bemowo? No. The district's active listings include a large share of new development projects (Chrzanów, the area around the M2 metro extension) priced at 17,000-19,000 zł/m², while the registered transactions of the past 12 months are dominated by older housing stock at 13,000-14,000. The asking and transaction medians partly describe different apartments. The gap is real as a description of the market ("new supply is priced far more ambitiously than what has been selling") — but it is not a discount you can pocket.

3. Adverse selection: overpriced listings hang around the longest

The pool of active listings is not a neutral sample of the market. A well-priced apartment disappears from the portal in 2-3 weeks. An apartment overpriced by 15% hangs for months — inflating the asking median every single day.

You can see it directly in the data: the median time on market for active Warsaw listings is 27 days, but one in five listings (18%) has been up for over 60 days, and one in eleven for over 90. That long tail consists largely of exactly those prices detached from the market. The longer a listing hangs, the more it distorts asking-price statistics — and the louder the negotiating signal for you.

4. The seller's psychological anchor

"I bought it in 2021 for 720k, put 80k into renovations, I need to get 950." The market doesn't reimburse costs — but try explaining that to someone who's spent three years living inside their "smart investment". Some listings aren't a negotiating strategy at all, just an emotional valuation that will never meet a single buyer. They can hang for six months and end with the listing being pulled rather than sold — while spending those six months distorting the median.

💡 Maybe prices are simply rising, and that's the whole gap? Not in this cycle. Warsaw's transaction median has been moving in a band of 14,800-15,900 zł/m² for over a year, with no clear trend. The 2026 asking-to-transaction gap isn't "delayed growth" — it's the combination of the four mechanisms above. In a market growing 15% a year, the same table would look completely different and would demand a different interpretation.

The Żoliborz case: how a median can lie

Back to the strangest row in the table. Żoliborz: transactions at 23,606 zł/m², listings at 20,370 zł/m². What happened here?

Żoliborz is a small district — 266 registered transactions in 12 months, 201 active listings. With samples this size, composition drives the median. And here the numbers speak for themselves: one in three transactions over the past year (91 of 266) was in a building from 2022 or later, with a median of 25,500 zł/m² — handovers of expensive apartments in new developments, whose deeds get registered in bulk once construction completes. Among active listings, buildings from 2022+ make up barely 3% (6 of 201) — the supply is dominated by older stock that owners list at "ordinary" Żoliborz prices around 20,000 zł/m².

The takeaway is universal and worth memorising: a district median — asking or transaction — is background, not a valuation of your apartment. The median doesn't know that your candidate is a third-floor walk-up in a prefab block, while half of this year's transactions nearby were luxury handovers. You have to compare like with like: same building, same street, similar construction year, size and condition — and only on that sample judge whether the asking price is fair.

That's exactly how mieszk.pl works: paste an Otodom listing link, and the system matches it against RCN transactions from the same building, street and micro-area (adjusted for floor, construction year and size), instead of comparing you to the median of an entire district. The difference between "this listing is 9% above the Mokotów median" and "three apartments in this very building sold in the past year at 16,200-16,800 zł/m², and this one is asking 18,500" — is the difference between a fun fact and a negotiating argument.

How much can you actually negotiate off in 2026?

If the median gap isn't a "discount waiting to be claimed", how much does actually come off asking prices? Watching transactions and listing behaviour in Warsaw in 2026, a fairly consistent pattern emerges:

• Fresh, market-priced listing (will be gone in 2-3 weeks): negotiating room 0-3%. Sometimes zero — good apartments at good prices can attract several bidders.
• Listing up for 30-60 days: a standard 3-7%. The seller already knows they overshot but isn't panicking yet.
• Listing up 90+ days, after price cuts: 7-12%, sometimes more if time pressure kicks in (buying another property, paying off a loan, dividing assets after divorce).
• Primary market: developers rarely cut the nominal price (it ruins their sales statistics and the pricing of future phases), but they do negotiate packages: parking space, storage room, finishing, payment schedule. Real value: 2-5% of the price.

Two things strengthen every one of these scenarios: knowing the listing's time on market (how long it has really been up — including its price-cut history) and comparable transactions from the immediate area. The first tells you how soft the seller has become. The second tells you where the market is. Without those two data points, negotiating is bidding blindfolded.

How to get at transaction prices yourself

RCN data is formally public — you just have to reach it. Your options, cheapest first:

1. Aggregate NBP and GUS reports — free, quarterly, but aggregated to city level. Good for sensing the trend, useless for a specific apartment.
2. A formal RCN extract from the county office / city hall — anyone can order a list of transactions for a specified area. Cost: usually a few dozen złoty per data set; waiting time: from a few days to 2-3 weeks; format: varies (sometimes a scanned table). Works if you buy once a decade and have patience.
3. A licensed appraiser's valuation report (operat szacunkowy) — 1,500-2,500 zł. Appraisers have RCN access and will produce a professional comparison. Sensible for unusual properties; for a standard two-bedroom it's a cannon aimed at a sparrow.
4. mieszk.pl — paste an Otodom link and in ~90 seconds get a comparison against real RCN transactions nearby (a 7M+ record database), the listing's time on market, its price-cut history, and a verdict on whether the asking price is in the market or above it. The fastest route when you're comparing several listings rather than one.

FAQ

If asking prices are "wishes", why look at them at all? Because they show supply and sentiment right now, without the registry's multi-month lag. A widening asking-to-transaction gap in a district is often the first signal that sellers are running ahead of the market (or that expensive new supply is arriving — as in Bemowo). The most informative thing isn't either number alone, but precisely the gap between the two price lists and how it changes over time.

Why does RCN data lag? A notarial deed travels from the notary to the county office, gets entered into the registry there, and only then reaches databases. In practice, the full set of transactions for a given month becomes visible after 3-6 months. That's why serious analyses base the transaction median on complete, "closed" months — recent ones are incomplete and can still shift.

Is the asking-to-transaction gap stable over time? No. It widens when the market slows (sellers refuse to accept the new reality) and narrows when the market runs hot (apartments sell at full asking, sometimes above). The current ~7-8% in a typical Warsaw district is the portrait of a balanced market with a slight buyer's edge.

Is the district's transaction median a good negotiating anchor? A weak one. The seller will bat it away easily: "mine is freshly renovated / has a garden / is next to the metro". The strong anchor is transactions from the same building or street in similar condition — there's no good comeback to that, because it's the same product, not a statistic.

---

The gap between asking and transaction prices in Warsaw in 2026 is a real 0.5-20% depending on the district — but as a negotiating tool it only works if you understand what it's made of: how much is genuine negotiating margin, how much is a different supply mix, and how much is overpriced listings that won't sell anyway. The median opens the conversation. Transactions from the specific building close it.

Before you make an offer, check on mieszk.pl how the listing price compares with actual transactions in the immediate area — and negotiate with numbers from notarial deeds, not numbers from other people's wishes.