Two methods, and we tell you which one you got

Most valuation sites give you one number from a model they will not describe. When it is wrong you have no way of telling why. This site uses two clearly different methods, says which one produced your figure, and shows the arithmetic either way.

Method one: the repeat-sales index

This is the good one, and you can use it if you know what you paid and roughly when.

The Federal Housing Finance Agency publishes a house price index built from repeat sales. Rather than comparing the average price of whatever sold this quarter against last quarter, which would move simply because a few expensive houses changed hands, it tracks properties that have sold more than once and measures how much the price of the same house changed between sales. That is the entire point of the construction: it holds the house constant to isolate market movement.

So if you paid $450,000 in 2019 and your metro's index has risen 41% since, the indexed value of your home is about $634,000. This is not a workaround or an approximation of a better method. It is what the index exists to do.

We hold this index for all 50 states and 410 metropolitan areas, currently to 2026Q1. Choosing your metro rather than leaving it statewide materially improves the answer, because metros inside a state routinely diverge by several points a year.

What it still cannot see

The index describes an area. It does not know that you replaced the roof, or that the house has not been touched since 1998, or that a new development went up at the end of the road. Over a long holding period those differences compound, which is why the range we show widens the longer you have owned the property.

Method two: the area benchmark

If you do not know the purchase price, we fall back to something much rougher. We take the national median sale price, convert it to a price per square foot, scale it by how your area's index compares with the typical US state, and multiply by your living area.

This is a market bracket, not a valuation. It knows nothing about your street, your lot, your view, or your school district, and any one of those can move a real sale price by more than the range we quote. We show it because it is better than nothing and because it is honestly labelled, not because it is good.

If you can find your purchase price, use method one. It is worth the two minutes.

What we do not have

• No property records. No deeds, no parcel data, no assessor history. We cannot look up your address.
• No comparable sales. Recent sales on your street are the strongest single input to a real valuation. That data is licensed and we have not licensed it yet.
• No photographs and no inspection. Condition is whatever you tell us it is.

When we do license property data, the address-level estimate becomes possible and this page will change to describe it. Until then, saying so plainly is more useful than a confident number with nothing behind it.

What we are certain about

The calculators are exact. Mortgage payments, refinance break-even, total interest, equity and borrowing limits are arithmetic on numbers you supply, and they are simply correct. Only the valuation is a model.

Full sources and the exact figures behind every adjustment are on the methodology page.