Australian Postcodes & Suburbs

A clean, machine-readable list of every Australian postcode + suburb / locality.

• 16,511 rows across all 8 states and territories
• Updated 2026-05-27
• Available as CSV, TSV, JSON, and Markdown — at the full level, per-state, or by postcode prefix
• Approximate Lat/Lon carried over from the 2021 snapshot for the 15,599 rows where the (postcode, suburb, state) key still matches; 912 rows new in 2026 have empty coordinates
• MIT-spirit "as-is" licence — see the bottom of this file

Quick links

I want…	File
Everything, with coordinates and category	australian-postcodes.csv
Everything as TSV	australian-postcodes.tsv
Everything as JSON	australian-postcodes.json
Just one state	data/by-state/<STATE>.csv (csv / json / md)
Just suburb→postcode (no coords)	data/lookup/postcodes-lookup.csv
Just suburb→postcode for one state	data/lookup/by-state/<STATE>.csv
Lookup by postcode region	data/by-postcode-prefix/<0-9>.csv
Today's snapshot, immutably dated	australian-postcodes-2026-05-27.csv
2021 historical snapshot	australian-postcodes-2021-04-23.csv

See data/README.md for the full directory map.

Schema (main files)

Column	Type	Example	Notes
Postcode	string	2000	4 digits, leading zeros preserved (0810, 0200)
Suburb	string	SYDNEY	Suburb / locality name, UPPERCASE
State	enum	NSW	One of ACT, NSW, NT, QLD, SA, TAS, VIC, WA
Lat	string	-33.860	Approximate latitude (3 dp), empty for some new rows
Lon	string	151.210	Approximate longitude (3 dp), empty for some new rows
Category	enum	Delivery Area	Delivery Area or Post Office Boxes

Postcode is a string, not an integer. Australian postcodes have leading zeros (e.g. 0810 for Darwin suburbs). If you load this into pandas, pass dtype={'Postcode': str} or use parse_options=… string.

Row counts

State	Rows
NSW	4,836
QLD	3,423
VIC	3,187
SA	1,946
WA	1,845
TAS	784
NT	339
ACT	151
Total	16,511

Of those, 16,237 are Delivery Area (actual residential / business localities) and 274 are Post Office Boxes ranges. Filter on Category == "Delivery Area" if you only want real suburbs.

Examples

CSV

Postcode,Suburb,State,Lat,Lon,Category
2601,ACTON,ACT,-35.280,149.110,Delivery Area
2602,AINSLIE,ACT,-35.260,149.150,Delivery Area
2914,AMAROO,ACT,-35.170,149.130,Delivery Area
2000,SYDNEY,NSW,-33.860,151.210,Delivery Area
3000,MELBOURNE,VIC,-37.810,144.960,Delivery Area
4000,BRISBANE,QLD,-27.470,153.020,Delivery Area

JSON

[
  {"postcode":"2000","suburb":"SYDNEY","state":"NSW","lat":"-33.860","lon":"151.210","category":"Delivery Area"},
  {"postcode":"3000","suburb":"MELBOURNE","state":"VIC","lat":"-37.810","lon":"144.960","category":"Delivery Area"}
]

Python (pandas)

import pandas as pd
df = pd.read_csv("australian-postcodes.csv", dtype={"Postcode": str})
df[df.State == "VIC"].head()

Python (per-state, minimum memory)

import pandas as pd
vic = pd.read_csv("data/by-state/VIC.csv", dtype={"Postcode": str})

JavaScript / TypeScript

import data from "./data/by-state/NSW.json" assert { type: "json" };
const sydney = data.filter(r => r.suburb === "SYDNEY");

Shell (suburb → postcode)

grep -F ',MELBOURNE,VIC,' australian-postcodes.csv
# or, much smaller file:
grep -F ',MELBOURNE' data/lookup/by-state/VIC.csv

AI-friendly slices

If you are loading this data into an LLM prompt, prefer the smallest slice that covers your query — every byte you avoid is tokens you avoid:

Use case	Recommended file	Approx size
Whole country, full schema	australian-postcodes.csv	822 KB
Whole country, no coords/category	data/lookup/postcodes-lookup.csv	339 KB
One state, full schema	data/by-state/<STATE>.csv	7 KB – 244 KB
One state, just suburb+postcode	data/lookup/by-state/<STATE>.csv	2 KB – 82 KB
Postcode region (e.g. all 3xxx)	data/by-postcode-prefix/<0-9>.csv	0.3 KB – 247 KB
Markdown table for inline reading	data/by-state/<STATE>.md	8 KB – 278 KB

ACT alone is ~150 rows and 2 KB — vastly cheaper to load than the full 16k-row file.

Libraries

Drop-in libraries for the four most common stacks, each with the dataset bundled (no network calls, no external dependencies). All live under libraries/ and share the same API surface.

Stack	Path	Install
Ruby (gem)	libraries/ruby	gem install australian_postcodes
Node (npm)	libraries/javascript	npm install australian-postcodes
Python (pip)	libraries/python	pip install australian-postcodes
Go (CLI+pkg)	libraries/go	go install github.com/schappim/australian-postcodes/libraries/go@latest
Homebrew CLI	tap	brew install schappim/auspostcodes/australian-postcodes

Every library exposes the same four operations:

• find_by_postcode(postcode) — every record matching a postcode
• find_by_suburb(suburb, state=...) — every record matching a suburb (case-insensitive), optionally narrowed by state
• postcode_for(suburb, state) — the postcode for a (suburb, state) pair, or null/nil/""/None
• all_in_state(state) — every record in a state

Ruby

Gemfile:

# Published gem
gem "australian_postcodes"

# Or straight from this repo (pin a tag/branch/commit)
gem "australian_postcodes",
    git: "https://github.com/schappim/australian-postcodes.git",
    glob: "libraries/ruby/*.gemspec"

# Or a local checkout
gem "australian_postcodes", path: "../australian-postcodes/libraries/ruby"

require "australian_postcodes"
AustralianPostcodes.postcode_for("Sydney", "NSW")  # => "2000"
AustralianPostcodes.find_by_suburb("Springfield", state: "QLD")

JavaScript / TypeScript

npm install australian-postcodes

const ap = require("australian-postcodes");
ap.postcodeFor("Sydney", "NSW");        // "2000"
ap.findBySuburb("Springfield", { state: "QLD" });

TypeScript types are bundled.

Python

pip install australian-postcodes

import australian_postcodes as ap
ap.postcode_for("Sydney", "NSW")        # '2000'
ap.find_by_suburb("Springfield", state="QLD")

A CLI is also installed:

australian-postcodes postcode-for Melbourne VIC      # -> 3000
australian-postcodes by-suburb Sydney --state NSW

Go

go install github.com/schappim/australian-postcodes/libraries/go@latest

import "github.com/schappim/australian-postcodes/libraries/go/postcodes"

postcodes.PostcodeFor("Sydney", "NSW")  // "2000"
postcodes.FindBySuburb("Springfield", "QLD")

The Go package compiles the dataset into a single postcodes.go source file (no go:embed, no separate .csv) — the resulting binary is ~3.8 MB and needs nothing else at runtime.

Homebrew (prebuilt CLI)

For macOS or Linux, the Go CLI is published as a Homebrew bottle via the schappim/auspostcodes tap. No Go toolchain required.

brew tap schappim/auspostcodes
brew install australian-postcodes

# or, one-shot
brew install schappim/auspostcodes/australian-postcodes

australian-postcodes postcode-for Melbourne VIC      # -> 3000
australian-postcodes by-suburb --state NSW Sydney

Supported architectures: macOS arm64 / amd64, Linux arm64 / amd64. Prebuilt tarballs are attached to each release.

Contributions

Pull requests welcome — particularly for:

• Filling in missing Lat/Lon on the 912 rows new in 2026
• Adding alternative slices (by-postcode-2digit, by-LGA, …) that consumers ask for
• Cleaning up suburb casing if your downstream prefers Title Case

When refreshing for a future release, please drop a dated snapshot (australian-postcodes-YYYY-MM-DD.csv) so prior snapshots stay reproducible.

Snapshots

• 2026-05-27 — Current dataset (australian-postcodes-2026-05-27.*)
• 2021-04-23 — Earlier snapshot, preserved as australian-postcodes-2021-04-23.*
• Lat/Lon coordinates are carried over from the 2021 dataset, matched by (postcode, suburb, state) key

License

THE DATA IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE DATA OR THE USE OR OTHER DEALINGS IN THE DATA.