U.S. ZIP+4 Database

Each edition offers different levels of geocoding precision and data depth:

• Standard ($399.95/yr): 37 ZIP+4 fields including address ranges, carrier routes, and legacy from/to lat/lon coordinates (~60% coverage, not actively maintained). Includes Standard ZIP Code Database. Best for basic address validation where geocoding precision isn't critical.
• Premium ($799.95/yr): 40 fields with street-level centroid coordinates (~80% coverage, actively maintained) and Census Block Group data. Includes Deluxe ZIP Code Database with demographics. Ideal for delivery optimization, territory mapping, and most business applications.
• Deluxe ($2,500/yr): 40 fields with building-level centroid coordinates (~95% coverage, highest precision available) and Census Block Group data. Includes Deluxe ZIP Code Database. Best for enterprise delivery systems, last-mile optimization, and mission-critical accuracy requirements.

View detailed field comparison →

The USPS assigns ZIP+4 codes efficiently based on mail delivery patterns:

• Residential streets: Typically one ZIP+4 covers all houses on a block side (e.g., 100-198 Main St all share ZIP+4 12345-1001)
• Large businesses: May receive a range of ZIP+4 codes for internal mail routing to different departments (e.g., PO Box 1000 might use ZIP+4 range 1000-1099)
• PO Box ranges: ZIP+4 often matches box numbers (PO Box 1004 = ZIP+4 1004), or a single box might have a full range

Because the USPS allows these range assignments, all ZIP+4 data must be stored in low/high range format per official standards - even when a range contains just a single code.

Our database provides the complete range information exactly as published by the USPS, ensuring USPS-compliant address validation.

The ZIP+4 database focuses purely on address ranges and delivery routes, so it doesn't include city names, time zones, or area codes. You must match it with the included 5-digit ZIP Code Database to get this essential information:

• Standard ZIP+4 includes: Standard ZIP Code Database (31 fields) with basic city, state, county, coordinates, and timezone data
• Premium & Deluxe ZIP+4 include: Deluxe ZIP Code Database (60 fields) with full demographics, statistical areas, and enhanced location data

How to match the databases:

• Use the PreferredLastLineKey field from ZIP+4 database
• Match it to CityStateKey field in the ZIP Code database
• This provides city name, state, time zone, area code, and all other location attributes
• Puerto Rico: Also match UrbanizationKey for urbanization names

See technical documentation with SQL examples →

Edition upgrades are pro-rated and applied instantly — pay only the gap between tiers, and your existing 12-month update window carries forward unchanged.

What actually changes on upgrade:

• Standard → Premium ($400 gap): Coverage jumps from legacy from/to endpoints to street-level centroids actively maintained across each USPS publication cycle. Census Block Group data refreshes with each release.
• Standard → Deluxe ($2,100 gap): Coverage moves to building-level centroids — the highest precision tier we publish, suitable for rooftop-grade and last-mile workflows.
• Premium → Deluxe ($1,700 gap): Same field count, but coordinates shift from street centerline to physical structure positions.

Process the upgrade directly from your account dashboard; enhanced files become downloadable immediately and your renewal date is preserved. Have a custom timeline, multi-product bundle, or volume requirement? Reach out — we'll handle it manually.

CSV (Comma-Separated Values) format only. With 48.5 million records, this is the most efficient format for handling large datasets:

• Why CSV only: Excel and Access have row limitations and can't handle files this large. CSV provides the best balance of file size, compatibility, and import speed.
• Universal compatibility: Works with all database systems (MySQL, PostgreSQL, SQL Server, Oracle, etc.) and programming languages (Python, PHP, Node.js, Java, C#, Ruby, etc.)
• 62 state/territory files: Data is broken into manageable state-level files for easier imports and faster processing
• Total size: >1GB compressed across all files - see breakdown by state

Most databases support direct CSV import commands like LOAD DATA INFILE (MySQL) or BULK INSERT (SQL Server), making integration straightforward.

The ZIP+4 database contains address ranges and delivery information but lacks city names, time zones, and demographic data. Here's how to access that information:

Basic matching process:

1. Import both the ZIP+4 database and the included ZIP Code database
2. Join using: ZIP+4.PreferredLastLineKey = ZIPCode.CityStateKey
3. This provides city name, state, time zone, area code, demographics, and all other location data

Puerto Rico special handling:

• Also match ZIP+4.UrbanizationKey = ZIPCode.CityStateKey to get urbanization names
• Use the same process for municipality names

Why not just use ZIP Code? While most people match by 5-digit ZIP Code, using PreferredLastLineKey is more precise and provides the exact city designation for that specific ZIP+4 record.

See SQL examples and detailed matching logic →

Standard Edition limitations:

• From/To lat/lon coordinates are ~60% coverage and not updated
• Census tract/block data is outdated
• Suitable only for basic address validation where precise location isn't critical

Premium Edition advantages:

• Street-level centroid coordinates: ~80-85% coverage, actively maintained
• Census Block Group data: Current 12-digit codes for demographic analysis
• Best for: Delivery route optimization, territory mapping, distance calculations, service area analysis
• Includes Deluxe ZIP Code Database with full demographics

Deluxe Edition advantages:

• Building-level centroid coordinates: ~95% coverage, maximum precision
• Centered on actual structures: Not street centerlines, but physical building locations
• Best for: Enterprise delivery systems, last-mile optimization, precise address validation, GIS applications, mission-critical accuracy
• Industry-standard quality: USPS-licensed ZIP+4 geocoding with monthly updates

Compare all editions →

"ND" stands for "Non-Deliverable." The USPS includes these addresses in their system but does not actively deliver mail to them.

Common examples of non-deliverable addresses:

• Vacant lots
• Land bordering railroad tracks
• Undeveloped parcels
• Areas designated as non-residential

Official USPS guidance: "Non-deliverable areas are areas to which the USPS does not deliver mail, such as vacant lots and land that borders railroad tracks. Mailers should not match an address identified as a non-delivery area."

If you encounter an ND designation during address validation, the address should be flagged as invalid for mail delivery purposes.

Yes! Complete worldwide coverage of all military addresses.

• APO (Army Post Office / Air Force Post Office): AE region (Europe/Middle East/Africa/Canada) - ZIP codes 09XXX
• FPO (Fleet Post Office): Navy, Marine Corps, Coast Guard - AP region (Pacific) - ZIP codes 962XX through 966XX
• DPO (Diplomatic Post Office): Department of State - AA region (Americas excluding domestic/Canada) - ZIP codes 340XX

All military installations, embassies, and overseas locations are included in the database with complete ZIP+4 coverage where assigned by the USPS.

Our database is used by the Department of Defense and all branches of the U.S. military for accurate mail routing worldwide.

Total compressed size: >1GB across 62 state and territory files (one file per state/territory).

Individual file sizes vary significantly:

• Smallest: 2-3 KB for small territories (Marshall Islands, Palau)
• Largest: 80-108 MB for populous states (Texas: 108 MB, California: 98 MB, Florida: 75 MB)
• Average: Most states range from 5-50 MB compressed

Yes - you can download only the states you need! The data is organized by state, so you can:

• Download just the states relevant to your business operations
• Import specific states to reduce database size
• Add additional states later as needed

See complete file size breakdown for all states →

Every coordinate in the ZIP+4 dataset is encoded in WGS84 (EPSG:4326), the geodetic reference used by GPS receivers, OpenStreetMap, Mapbox, ArcGIS, and the major web-mapping APIs. Drop our lat/lon values into any modern GIS pipeline without reprojection.

What differs between editions is the geocoding semantics, not the projection:

• Standard: Encodes two endpoints per record — the address range's first and last delivery points. Legacy AIS-derived data, ~60% populated, frozen at the geocoding state from when it was first published.
• Premium: One centroid per record, snapped to the street centerline. ~80% populated, refreshed monthly against the USPS publication cycle, ideal for territory drawing and ETA modeling.
• Deluxe: One centroid per record, snapped to the actual structure (rooftop). ~95% populated and rebuilt continuously. Suited for last-mile dispatch, drone-delivery test corridors, and any workflow that can't tolerate "near the street."

Output is in decimal-degree format at 6-place precision (~11 cm at the equator). Negative longitudes for the Western Hemisphere; no swapped-axis surprises.

Need full polygon coverage instead of point geocoding? Pair with our ZIP Code Boundary Data shapefiles.

Yes! The ZIP+4 Database is perfect for building USPS-compliant address validation systems.

What you can validate:

• Verify street addresses exist within their ZIP code
• Confirm address ranges (e.g., 123 Main St is within 100-198 range)
• Validate ZIP+4 assignments for accurate mail sorting
• Check carrier routes for delivery feasibility
• Identify non-deliverable addresses (ND designations)
• Standardize address formats per USPS guidelines

Common implementation patterns:

• Real-time validation: Check addresses at point of entry in forms/checkout
• Batch cleansing: Validate and standardize existing address databases
• Mail preparation: Assign correct ZIP+4 codes for USPS automation discounts
• API development: Build address validation services for your applications

Our data is USPS-licensed and updated monthly to ensure compliance with current postal standards. See implementation examples →

The database contains 48.5 million records representing approximately 65.4 million individual ZIP+4 codes.

Why the difference? Because ZIP+4 codes are stored in ranges:

• A single database record might represent one ZIP+4 (e.g., 12345-1001 to 12345-1001)
• Or it might represent a range of 100 codes (e.g., 12345-1000 to 12345-1099)
• The 48.5M records efficiently represent all ~65.4M possible ZIP+4 assignments

View complete technical specifications →

ZIP+4 data shifts faster than 5-digit ZIP data — roughly 5% of records turn over every month as the USPS adds new construction, retires vacated ranges, renumbers carrier routes, and re-walks delivery points. Every license includes 12 months of synchronized publication at no extra cost, with new data dropped on the 1st alongside the parent ZIP Code dataset.

Pull methods we provision per license:

• FTPS / SFTP credentials: A dedicated account on our publishing server. Schedule a cron job, Windows Task, or Airflow DAG that pulls overnight; we keep the prior month online for one full cycle for rollback.
• Browser download: Same files exposed in your account dashboard if your environment prefers manual handoff to ops.

Two artifacts ship every release:

• Full state-segmented CSVs: All 62 state and territory files at current-month coverage. Conventional swap-and-truncate pipelines rebuild the table from scratch, reindex, and swap the view/synonym.
• Transaction-only deltas: Two-stream changeset (deletes then adds) so customers running 48.5-million-record installs can apply diffs instead of full reloads when their pipeline is incremental.

Cadence advice: The 1st-of-month drop is firm; republications to correct a flagged anomaly happen rarely but do happen — almost always before the 5th. Schedule the automated ingest for the 5th to sidestep the rare race condition. Subscribers get an emailed manifest with record counts and change deltas for every release.

Once your 12 months wrap, you keep every CSV you downloaded; you only renew if you want the next monthly drop.

A 30-day grace window opens automatically when your subscription lapses — full data usage continues so renewal doesn't break a production pipeline. After day 30, our license terms require you to stop using the data and remove it from every production system, snapshot, backup, and derivative table within 10 business days.

Why ZIP+4 specifically punishes stale data more than 5-digit ZIP:

• ZIP+4 turns over at roughly 5% per month — an order of magnitude faster than 5-digit codes. A 2-year-old dataset is missing more than half the current carrier-route assignments.
• New residential construction inherits brand-new ZIP+4 ranges that weren't in the old file at all — your validator will keep rejecting every new-build address.
• Vacated ranges (demolished buildings, retired PO Boxes, abandoned apartment complexes) stay in your stale data forever. You'll route mail to addresses USPS no longer recognizes.
• Delivery-point status shifts as USPS re-walks routes — an address that was deliverable last cycle may not be today, or vice versa.
• Building-level centroids drift as buildings are surveyed or rebuilt; the rooftop coordinate from 2023 may point at the wrong structure after a teardown or replat.

What this looks like in your application:

• USPS-style validation rejects valid new addresses that the Postal Service added after your snapshot
• Carrier-route assignments fall out of sync, costing your pre-sort mailings their automation discount
• Delivery-point geocoding misses recently constructed homes entirely — "address not found" at checkout for legitimate customers
• Mail returned as undeliverable for ranges that have since been vacated, with no signal in your data that the range no longer exists

Industry data-decay benchmarks (Gartner, MarketingSherpa): 22% of operational address data becomes stale within a year, B2B contact files erode at 2.1% per month, and poor data quality is estimated to cost the average enterprise roughly $15M / year — numbers that disproportionately load onto teams sitting on a frozen ZIP+4 file.

Pricing tip: lock 2- or 3-year renewals at 10% off via the purchase options — cheaper than rebuilding accuracy after a long gap.