Method
How this was built, and what it cannot tell you
Everything here comes from public records. This page says exactly which records, exactly what we changed, and where the numbers stop being reliable.
The limitation that matters most
Around 80% of New York City dogs are not licensed. This is not a register of every dog in the city. It is a register of the dogs whose owners filled in a form, and those owners are not a random sample of dog owners.
Every figure on this site is a rate per 10,000 licensed dogs, either within a birth-year group or within a ZIP code. That controls for the obvious problem: a neighborhood that licenses at half the rate of another is not thereby made to look as though it has half the dogs. What it cannot control for is whether the sort of person who licenses a dog in one neighborhood names dogs differently from the sort of person who does not. That caveat applies to every neighborhood comparison here and we would rather say so than let you find out later.
Source
| Dataset | NYC Dog Licensing Dataset |
|---|---|
| Publisher | NYC Department of Health and Mental Hygiene, via NYC Open Data |
| Retrieved | August 21, 2026 |
| Terms | NYC Administrative Code 23-502(d): no license requirement or restriction on use; source, version and modifications to be identified. |
| Records | 819,323 license rows, reduced to 302,213 individual dogs |
What we changed
- License renewals deduplicated to individual animals
- Names normalized for case and punctuation; placeholder names removed
- Breed free text reduced to a single primary breed
- Records outside the cohort range or without a valid ZIP excluded
The first one matters most. Licenses renew, so the same dog appears in the raw file once for every renewal. Counting rows rather than animals would make long-lived dogs count many times over and tilt the whole dataset toward older names, which is why 819,323 license records become 302,213 individual dogs here. A further several percent of records carry a placeholder where the name should be, and those are excluded rather than counted as names.
One thing we deliberately did not change. On some breed pages the most common entry is the breed word itself: 135 of the 6,296 licensed French Bulldogs are recorded as “French,” and 120 of the Havanese as “Havanese.” Nothing in a license record distinguishes a dog whose breed was written into the name field from a dog an owner really did name after its breed, and both happen. Deleting those rows would mean guessing, so they are kept, marked with a dagger on the breed page, and left out of the sentence that says what the breed is usually called. A name is only marked this way when it is a word of that breed’s own name and at least half the dogs carrying it are that breed, which is why Charles stays on the Cavalier King Charles Spaniel page: only 7% of the Charleses in the city are Cavaliers, so it is plainly a name someone chose.
Thresholds
| Name page | A name needs at least 10 dogs. Below that the trend line is noise. |
|---|---|
| Neighborhood | A ZIP code needs at least 1,500 licensed dogs before it is used in any comparison. |
| Breed page | A breed needs at least 150 dogs. |
| Years | Dogs born 2005 to 2023. Outside that range there are too few records to plot. |
Pup Census is an independent project. It is not affiliated with, endorsed by, or connected to the City of New York, the Department of Health and Mental Hygiene, NYC Open Data, or any other city, county or government agency. City and agency names are used descriptively to identify the source of the data.