Orlando's 51,112 Permits Sit on 17,932 Addresses

Orlando's building department has issued 51,112 permits since January 2025, and they sit on 17,932 distinct addresses. That is 2.85 permits for every property in the file. Anyone treating an Orlando permit count as a count of buildings, homes or projects is overstating the market by nearly three times.

The more useful finding is where that inflation actually lives. It is almost entirely cross-trade stacking and a small number of enormous campus addresses. Inside any single trade, one permit is very close to one job.

TL;DR

• Orlando: 51,112 permits on 17,932 addresses, a multiple of 2.85.
• 56.2% of addresses carry one permit; 85 addresses carry 10.6% of the file.
• Filter to your own subtype first and the multiple falls to 1.04–1.30.
• Count addresses for properties, permits for jobs. Decide which first.

How many permits does one Orlando property actually generate?

On average, 2.85. Over the trailing 12 months the figure is 2.54, across 29,418 permits on 11,586 addresses, so this is a stable property of the file rather than an artefact of one busy quarter.

The average hides more than it reveals. The distribution is deeply lopsided.

Permits at the address	Addresses	Share of addresses	Share of all permits
1	10,077	56.2%	19.7%
2	3,211	17.9%	12.6%
3	1,139	6.4%	6.7%
4–5	1,197	6.7%	10.3%
6–10	1,715	9.6%	26.2%
11–25	508	2.8%	14.0%
26+	85	0.5%	10.6%

Most Orlando addresses appear once. A majority of Orlando permits do not belong to those addresses.

The 85 properties in the bottom row are worth naming, because they are not houses. 5900 Universal Blvd, inside the Universal Orlando resort complex, carries 609 permits across six permit types. The Mall at Millenia at 4200 Conroy Rd carries 382. AdventHealth Orlando at 601 E Rollins St carries 358. These are campuses filing continuous work under one street address, and they behave nothing like the residential file around them.

Strip those 85 addresses out, along with the 90 rows that carry the literal placeholder value NO ADDRESS, and the multiple falls to 2.56 across the remaining 45,718 permits. That is the number to use if you are sizing the ordinary Orlando market.

Why does the 2.85 multiple mislead almost everyone?

Because it invites the wrong correction. The instinct on seeing 2.85 is to deduplicate every pull down to unique addresses, and for most contractors that is the wrong move.

Split the repetition into its two possible causes. An address can carry several permits because different trades each filed on the same job, or because the same trade came back. In Orlando, 34.4% of addresses carry more than one permit type, and only 9.4% repeat inside a single type.

So the file is stacked sideways, not vertically. The duplication is trades working next to each other, which is exactly the thing a single-trade filter removes for free.

That distinction matters commercially. A permit is a signal about the owner of the property, not about the contractor who filed it, and when six trades file on one lot you are looking at six views of a single owner decision. Deduplicate and you see the decision once. Leave the file alone and you see every trade seat that decision created.

Which Orlando trades should not deduplicate at all?

Most of them. Filter to a permit subtype and the multiple collapses toward one.

Subtype	Permits	Addresses	Permits per address
Plumbing - New	1,641	1,580	1.04
Mechanical Permit - New	1,621	1,563	1.04
Building Permit - Roof	6,128	5,690	1.08
Engineering - SFR	1,686	1,564	1.08
Building Permit - New	2,179	1,912	1.14
Electrical - New	2,099	1,838	1.14
Building Permit - Repair	3,037	2,658	1.14
Plumbing - Repair	2,169	1,837	1.18
Mechanical Permit - Repair	4,280	3,296	1.30
Building Permit - Alteration	2,621	1,766	1.48
Fire Permit - FireSupp	1,580	1,007	1.57
Electrical - Alt	2,459	1,528	1.61
Electrical - LowVoltage	3,578	2,156	1.66

A roofer pulling Orlando's 6,128 roof permits is looking at 5,690 distinct roofs. Deduplicating that list removes 438 rows and costs real work. The same goes for new plumbing and new mechanical, where the multiple is 1.04 and deduplication is close to a rounding error.

The trades at the bottom of the table are the exceptions. Low-voltage electrical at 1.66 and fire suppression at 1.57 genuinely do return to the same address, because both are commercial-weighted and large buildings file them in phases. If you sell into either, the correction is worth applying.

Note also that these subtype multiples are all far below 2.85. That is the whole argument in one line: the headline number is a property of the file, not of your market.

What does the stacking look like on a single lot?

Take 10233 Hadley St, a single address in the EverBe community in East Orlando. It carries seven permits, and every one of them is a different subtype: Building Permit - New, Building Permit - Roof, Electrical - New, Electrical - LowVoltage, Engineering - SFR, Mechanical Permit - New and Plumbing - New.

One house. Seven filings. Seven separate openings for seven different trades, and one owner decision behind all of them.

Nothing on that lot was permitted twice. This is what normal new construction looks like in the file, and it is why a naive permit count reads as seven properties when it should read as one.

Neighbouring lots on the same street show the pattern with different edges. 10218 Hadley St carries five permits and 10274 carries six, each drawing a slightly different mix depending on what the builder subbed out and when. The stack is consistent in shape and variable in depth.

How do you apply the correction factor?

Start by writing down which question you are answering. Properties or jobs. Almost every bad Orlando market estimate comes from answering one and reporting the other.

For a properties count, deduplicate to address and then decide what to do with the campuses. Removing the 85 heavy addresses is usually right for residential work and usually wrong for commercial, where those campuses are the market.

For a jobs count in your own trade, filter to your subtype and stop. The 2.85 multiple does not apply to you, and applying it will cut your addressable list by roughly two thirds for no reason. That error points in the expensive direction, and it is easy to make when reading the clustered filings inside Orlando's master-planned communities.

For anything in between, use 2.56 rather than 2.85. It excludes the campus distortion and the placeholder rows, and it is the closer estimate for ordinary Orlando property.

One more practical note. Orlando records carry no coordinates, no ZIP fields and no unit numbers, so the address string is the only property key this file has. That makes normalisation load-bearing. Two spellings of the same street become two properties, and your multiple drifts. It also means the difference between a raw permit list and a scrubbed lead list is doing more work in Orlando than it would in a geocoded market.

What will this method not tell you?

It will not tell you how many homes Orlando built. The file mixes residential and commercial without a unit count, and a deduplicated address is a property, not a dwelling. An apartment building and a single-family lot each count once.

It will not transfer. This multiple is derived from City of Orlando filings, and Orange County runs a separate permitting system for unincorporated areas that we do not hold. We cover 1 of Florida's 67 counties, and no correction factor derived here should be carried into another market.

It also revises something we published two days ago. That EverBe analysis found 6.38 permits per address inside three master-planned communities and generalised it to roughly six times citywide. The generalisation was wrong. Production master plans stack far harder than the rest of Orlando, and the citywide figure is 2.85. The community number stands; the extrapolation from it does not.

Finally, none of this speaks to timing or approval speed. Orlando's decision date mirrors its application date in this file, so turnaround cannot be measured from it, and we do not try. What the $7,500 exemption analysis showed about reading regulatory change out of filing volume applies here too: be careful which questions the columns can actually answer.

How permits.llc fits in

We hold the City of Orlando file with 95.6% contractor names and 98.5% owner names on every one of those 51,112 permits, updated to within about a day. The address-level detail in this article is not a separate product. It is what the raw records look like when you group them.

If you want to run these cuts yourself, the Orlando permit download gives you the underlying records, and the Orange County coverage page shows exactly what is in the file before you commit to anything. Contractors who want the same filings arriving daily rather than as a bulk pull can set up permit alerts instead.

Either way, decide whether you are counting properties or jobs before you count anything. In Orlando the two answers differ by 2.85, and only one of them is the one you meant.