How Much Will My House Be Worth in 10 Years? The 3 Percent Rule Is Off by Nearly Half

Ask ten homeowners what their house gains in value each year and most will land on the same number: about 3 percent. It shows up in retirement spreadsheets, rent-versus-buy debates, and the back of every napkin. So we checked it against what homes actually did.

We matched 6.96M repeat-sale pairs (matched consecutive sales of the same home), and across the 3,124 US cities (postal areas) with at least 300 pairs each, the measured median realized appreciation was 5.27 percent a year (exact 5.265, median of city medians). The folk 3 percent assumption was not just low. It was off by nearly half.

Compounding turns that gap into serious dollars. A $300,000 home growing at 3 percent reaches $403,175 in ten years. If the next ten years repeated the measured median rate, the same home would reach $501,142, with the middle half of cities spanning $468,125 to $541,323. That is a difference of $97,967, about 24.3 percent, from a 2.27-point difference in the assumed rate.

One thing this article will not do is forecast. Nobody can tell you what your house will be worth in 10 years, and anyone quoting a single confident number is guessing. What we can tell you, precisely, is what homes did. Then you can decide whether the next decade rhymes.

The 3 percent rule ran low in 99.6 percent of cities

Across the 3,124 qualifying cities (postal areas), 3,111, or 99.6 percent, delivered a median realized rate above 3 percent a year, and 1,843 cities, 59.0 percent, came in above 5 percent. Two caveats ride with that claim in the same breath: a repeat-sales measure only sees homes that sold at least twice, and cities below the 300-pair floor are invisible to this cut, so the sample skews toward larger postal cities.

The middle half of cities ranged from 4.55 to 6.08 percent a year (p25 to p75), with the 10th percentile at 4.01 percent and the 90th at 6.84 percent. Even a pessimistic city draw historically beat the folk rule.

The $300,000 question, at measured rates

The table below is illustrative compounding of a historical realized rate, in nominal future dollars. It answers "what if the future repeated the measured past," never "what will happen." Dollar figures are computed from the unrounded 5.265 percent.

At five years, the measured median lands at $387,741 inside a middle-half band of $374,750 to $402,985, a difference of $39,959 over the 3 percent figure. At twenty years the compounding gap goes vertical: $837,146 at the measured median, inside a middle-half band of $730,469 to $976,768 across cities, versus $541,833 at 3 percent, a difference of $295,313, or 54.5 percent more in nominal dollars.

And "nominal" is doing real work in that sentence. None of these rates are inflation-adjusted. At about 3 percent CPI, the measured 5.27 percent a year is roughly 2.2 percent a year in real purchasing power, and every dollar figure above is nominal future dollars, not present-day wealth. The folk 3 percent rule is itself a nominal rule of thumb, which means that in real terms it implies almost no gain in purchasing power at all.

What we measured

For every home in our sold corpus (houses, condos, townhomes) that closed at least twice, we computed the annualized price change between consecutive sales, on holds of 1 to 30 years, with no recency window. The data therefore includes the 2020-2022 run-up alongside older, slower regimes. Each city's figure is the median of those annualized changes, and the national figure is the median of city medians (unweighted), so tiny Bushkill, PA counts exactly as much as Miami, FL with its 31,847 pairs. A pooled-pair national median, weighting every pair equally, is a different number and is not derivable from this snapshot, which stores only per-city medians.

The corpus: 6,963,756 repeat-sale pairs (matched consecutive sales of the same home), snapshot dated 2026-08-14. The rankings cover the 5,014,557 pairs, about 72 percent of the total, that sit in the 3,124 cities (postal areas) with at least 300 pairs each. Twelve states are absent entirely, largely non-disclosure states: AK, ID, KS, LA, MO, MS, MT, ND, NM, TX, UT, WY. "National" here excludes them.

One reconciliation note: our published real home appreciation study reports a 5.19 percent national median under a stricter qualification gate covering 1,472 areas. This article uses the wider cut, 3,124 cities (postal areas) at 300-plus pairs, and lands at 5.27 percent. Both round to about 5.2 percent a year; the gates differ, so the numbers do too.

Your city matters far more than the national median

Realized rates across cities (postal areas) ran from 1.49 to 11.36 percent a year, a spread no single national assumption can paper over.

Top cities (postal areas) by realized annual appreciation:

Bottom cities (postal areas): Two reading notes. These are postal areas, not municipalities: Bushkill and Tobyhanna are Poconos vacation submarkets, and "New York, NY" is the Manhattan postal city only, not New York City. Brooklyn, NY is a separate key that delivered 4.71 percent a year across 8,082 pairs, so Manhattan's 1.49 percent likely reflects its condo and co-op heavy deed mix rather than a citywide story. A single city can also hide two very different markets inside it, a pattern we documented in our two-market cities study.

The state table

State figures are the median of city medians within each state, gated at five or more qualifying cities.

The bottom five: Iowa 3.96 percent (38 cities), Maryland 4.10 percent (97), Illinois 4.20 percent (176), Ohio 4.45 percent (163), Minnesota 4.47 percent (73). Maine's chart-topping 8.83 percent rests on just 10 cities and 5,066 pairs, so treat small-state medians as fragile.

So how much will your house be worth in 10 years?

Honestly: unknown, and we will not pretend otherwise. Past appreciation is not a promise, the measured window includes an unusually hot stretch, and repeat-sales data only sees homes that sold twice. What the data supports is a conditional statement: if the next ten years repeated the measured median, a $300,000 home would reach about $501,142 in nominal dollars, with the middle half of cities spanning $468,125 to $541,323, versus $403,175 under the folk 3 percent rule. The single biggest upgrade you can make to that estimate is swapping the national number for your own city's realized rate, which is exactly what our free appreciation calculator does.

Related studies

This article is part of a four-study series from the same closed-sale corpus: Where Flippers Buy Houses at Half Price, What a Fixer-Upper Really Sells For: The 20.5% As-Is Discount, Measured in 580 Cities, and The Fixer-Upper Pipeline Map: Distress-Heavy Cities Posted 3.5x the Flip Margins.