Metro size barely matters for rent yield — here's the data
There's a piece of conventional wisdom that gets drilled into every cash-flow-vs-appreciation debate on these forums: small metros are where the yield lives. Cleveland, not San Diego. I held that belief for years. Then I actually looked at the distribution, and it mostly isn't true.
I pulled rent-to-price yields across 700 US metros as of April 2026 and grouped them by metro size. (Rent yield = annual rent ÷ home value × 100. Higher = more rental income per dollar of purchase price.)
The numbers
- Top 50 metros: 5.60% median yield (range 2.61%–7.93%)
- Mid-size (51–200): 5.92% median yield (range 3.63%–9.63%)
- Small (200+): 5.87% median yield (range 2.97%–13.44%)
Thirty-two basis points separate the biggest bucket from the smallest. And the mid-size bucket actually edges out the small one at the median. The "buy small for cash flow" rule isn't a rule — it's barely a tendency.
The real story is the spread inside each bucket. Small metros range from 2.97% to 13.44% — a 10-point swing. There are small metros with worse yields than half the Top 50, and Top 50 metros with better yields than half the smalls. Picking the bucket doesn't pick the deal. The metro picks the deal.
The number worth staring at
13.44% — the highest rent yield in the whole dataset right now. It belongs to Meridian, MS (size rank 397, home value ~$121K, annual rent ~$16K).
Before anyone opens a new tab to look up Mississippi land prices: a 13% yield in a metro nobody is moving to is usually a yield trap. The math works on paper. The vacancy rate, tenant quality, resale liquidity, and depreciation curve don't. The hard part of screening isn't finding the highest number — it's knowing when the high number is real and when it's a warning sign.
The question I'm chewing on
If metro size doesn't predict yield, what does? Vacancy trends, rent-growth trajectory, job migration, household formation rate — or something else? My current guess is rent-growth trajectory matters more than any static yield number, but I haven't pinned it down. Curious what the people here who actually own in these markets think predicts yield best.
— David
Most Popular Reply
Right finding, pointed one level too high. You showed size is the wrong axis — but the metro is still a bucket. Run the same test inside one metro across its tracts and the spread is wider than the 32bps you found between size tiers. The variation that matters was never between metros; it's inside them. On your closing question — what predicts yield if size doesn't: the demand denominator under the rent, not the yield itself. Household formation, vacancy direction, absorption. Yield is a price-to-rent snapshot; it says nothing about whether the rent holds. Which is what Meridian is telling you. 13.44% isn't a small-metro artifact — it's a thin-denominator print, and you'll find the same trap in specific tracts inside top-50 metros where a high headline yield sits on a demand base that's quietly shrinking. Does your data let you go sub-metro? Curious whether the within-metro dispersion holds up the way the cross-metro one collapsed.