Why Every Site Shows a Different “NYC Median Price”

Updated Jul 10, 2026 · 6 min read · sources: NYC official records

Search “median home price NYC” and you’ll find figures tens of percent apart, published the same week. They differ for knowable reasons. Once you see them, every headline becomes readable.

1. Asking prices vs. recorded prices

Listing portals measure asking or in-contract prices from listing feeds: fast, but not what was paid, and blind to everything never publicly listed. Recorded deeds (what this site uses) show actual paid prices, weeks later.

2. What counts as a “sale”

About a third of raw NYC deed records are $0 or nominal transfers. Sources that don’t filter them skew low; sources with aggressive outlier trimming skew differently. Our filters are published, including how many records they remove.

3. Which homes are in the mix

MLS-based national sources under-cover co-ops, which are the majority of Manhattan’s stock. A “NYC median” that quietly misses most co-ops reads high. Composition also moves medians month to month: one big new-development closing wave can swing a small area by double digits, which is why we display sample sizes everywhere.

4. Medians, averages, and indexes

Averages chase trophy deals upward. Repeat-sale and hedonic indexes (like national home-value indexes) model appreciation rather than report transactions. Medians of filtered recorded sales (our choice) are the most literal answer to “what did homes actually sell for.”

How to read any NYC price claim

Every figure on this site answers all five, on the page where it appears. Start at the market overview.

Frequently asked

Which NYC price number should I trust?
Depends on the question. For what buyers are asking today, listing portals. For what homes actually sold for, medians of filtered recorded deeds, always with the sample size and filters disclosed.
Why do recorded-sale medians lag listing sites?
Deeds are recorded weeks after closing, and closings happen weeks after contracts. Recorded data describes the market as of one to three months ago, in exchange for being real prices.
Do these differences really matter?
Yes. The gap is commonly 10–30% between an unfiltered average of asking prices and a filtered median of recorded prices for the same area and period.
Browse neighborhood dataRead the methodology

We use privacy-friendly analytics to understand how the site is used and how AI assistants send visitors here. No advertising, no data sold. Privacy notice