Road crashes, New South Wales
liveNSW Crash Data from Transport for NSW, one row per crash for the latest five reporting years. Each crash has its degree, year, month, day of week and two-hour band, street, town, council area and coordinates, the road and lighting conditions, the road user movement and definitions for crash type codes, the vehicles involved and the number of people killed and injured by severity.
Also called: NSW Crash Data, NSW road toll and crash statistics, Transport for NSW road crash data by crash, NSW car accident statistics.
- Which crashes are included?
- The crashes reported to NSW Police on a public road in the five reporting years of the release, in which a person was killed or injured or a vehicle was towed away. Each row is one crash, with its location, road and conditions and the number of people killed or injured.
- Why does each version cover only five years?
- Transport for NSW publishes a rolling five years each October and adds the release as a new file. This entry follows the newest file, so each version is one release. Earlier releases stay here as earlier versions.
- Can I count the road toll from this?
- The killed column counts people who died within 30 days of the crash, as the publisher reports them. Add it across the crashes of a year to get the deaths the publisher counted for that year.

Made for agents
Connect your AI agent
Any agent that connects to publicdata.au/mcp can find this dataset, read its 62 fields, count and filter across all 93,042 rows and diff its versions. Every answer names the version and carries TfNSW's attribution. No key.
Then ask it: What does Road crashes, New South Wales hold, and what changed in the newest version?
claude mcp add --transport http publicdata https://publicdata.au/mcp One command. No key, no account, no sign-up.
Settings, then Connectors, then Add custom connector, and paste https://publicdata.au/mcp.
On the web and in the desktop app.
Open https://chatgpt.com/plugins/plugin_asdk_app_6abbc4cec1a08191b61eaccfe2091100 and connect the publicdata.au plugin.
Listed in ChatGPT's plugin directory. publicdata.au asks for no key and no account.
https://publicdata.au/mcp Streamable HTTP. Cursor, VS Code or any client that connects to a remote MCP server.
Using https://publicdata.au/llms.txt, how many people died on Queensland roads in 2025, by month? For an assistant that reads the web and has no connector.
The dashboard above, live. Click to filter, add panels, share a link or embed a frame.
Query APIFilter and count from a URLaggregate?group=speed_limit&metric=count°ree_of_crash=eq.Fatal
Crashes by speed limit where degree of crash is Fatal: 452 100 km/h, 260 50 km/h, 245 60 km/h, 235 80 km/h.
Download it as CSV, Parquet, SQLite and 5 more formats
Pick a format and a version. CSV is picked first because it opens in the tools most offices have. The URL is yours to keep. A dated version keeps its content.
UTF-8 without a byte-order mark, RFC 4180 quoting, one header row using the normalised field names. Opens in Excel, Numbers and Sheets.
Columnar and typed. The smallest download and the fastest to query. Reads directly in DuckDB, pandas, Polars, Arrow and Spark. The provenance header is in the file metadata.
One database with a records table, a fields table and a publicdata metadata table. Open it with any SQLite client.
One DuckDB database with a records table of typed columns, a fields table and a publicdata metadata table. It attaches read-only over HTTPS, so DuckDB, Python and R can query it without downloading it.
A GeoPackage with one point layer, records, in the publisher's coordinate system, plus the fields and publicdata tables. Drag it into QGIS or ArcGIS.
GeoParquet: the Parquet table with a WKB geometry column and its coordinate system in the file's geo metadata. Reads directly in DuckDB, GeoPandas, QGIS and GDAL.
The first line is the provenance header, then one JSON object per line. For streaming and for tools that read line by line.
The same CSV, gzipped, at about a tenth of the size. curl, pandas, DuckDB and R read it directly.
One JSON file per value of year_of_crash, 6 files. The example is by/year_of_crash/2025.json. Every file carries the same header and a partition note.
Excel is not offered because the table is 58.1 MB as CSV, over the 50 MB limit for Excel. JSON is not offered because the table is 158.8 MB as NDJSON, over the 50 MB limit for JSON. GeoJSON is not offered because the file would be 168.0 MB, over the 100 MB limit for GeoJSON.
In your own tools
Use it in Excel, R, Python and more
Excel and Power BI read the CSV from its address and refresh from it. The R and Python packages take any dataset on this site by its slug, so a new dataset needs no new release. The DuckDB file attaches read-only over HTTPS, and a query reads only the blocks it touches.
The dated URL in the code never changes. https://publicdata.au/d/nsw-road-crashes/latest/ redirects to the newest version.
https://publicdata.au/d/nsw-road-crashes/latest/data.csv In Excel choose Data, then From Web, and paste this address. Excel keeps it, so Refresh All reads the newest version. A sheet holds about a million rows; past that, load the query to the Data Model.
https://publicdata.au/d/nsw-road-crashes/latest/data.csv In Power BI Desktop choose Get data, then Web, paste this address and choose Anonymous when asked how to sign in. A scheduled refresh reads the same address, so the report follows each new version.
library(publicdataau)
df <- pd_read("nsw-road-crashes")
pd_attribution(df) install.packages("publicdataau"). pd_read() fetches the version's Parquet file; pd_rows() and pd_aggregate() ask the query API instead, and pd_connect() attaches the DuckDB file.
import publicdata_au as pd_au
df = pd_au.read("nsw-road-crashes")
df.attrs["publicdata"]["attribution"] pip install "publicdata-au[pandas]". read() fetches the version's Parquet file; rows() and aggregate() ask the query API, and connect() attaches the DuckDB file.
INSTALL httpfs; LOAD httpfs;
ATTACH 'https://publicdata.au/d/nsw-road-crashes/v/2026-10-09/data.duckdb' AS nsw_road_crashes (READ_ONLY);
SELECT year_of_crash, count(*) FROM nsw_road_crashes.records GROUP BY 1 ORDER BY 2 DESC; The DuckDB file attaches read-only over HTTPS and only the blocks a query touches are read. Parquet works the same way: FROM read_parquet(url).
const res = await fetch("https://publicdata.au/api/v1/datasets/nsw-road-crashes/aggregate?group=speed_limit&metric=count°ree_of_crash=eq.Fatal");
const { rows, publicdata } = await res.json();
console.log(rows, publicdata.attribution); The query API answers a page on any site as well as Node, with no key. It returns the rows and the attribution the licence asks for.
What is in it
Column names are made snake_case and the publisher's original header is kept beside each one. Blank cells are null. Nothing is added, removed, ranked or summarised.
| Field | Type | Publisher's header | Note |
|---|---|---|---|
| crash_id | integer | Crash ID | |
| degree_of_crash | string | Degree of crash | |
| degree_of_crash_detailed | string | Degree of crash - detailed | |
| reporting_year | integer | Reporting year | |
| year_of_crash | integer | Year of crash | |
| month_of_crash | string | Month of crash | |
| day_of_week_of_crash | string | Day of week of crash | |
| two_hour_intervals | string | Two-hour intervals | |
| street_of_crash | string | Street of crash | |
| street_type | string | Street type | |
| distance | integer | Distance | |
| direction | string | Direction | |
| identifying_feature | string | Identifying feature | |
| identifying_feature_type | string | Identifying feature type | |
| town | string | Town | |
| route_no | integer | Route no. | |
| school_zone_location | string | School zone location | |
| school_zone_active | string | School zone active | |
| type_of_location | string | Type of location | |
| latitude | number | Latitude | As published. The publisher's data manual does not state the datum. |
| longitude | number | Longitude | As published. The publisher's data manual does not state the datum. |
| lga | string | LGA | |
| urbanisation | string | Urbanisation | |
| conurbation_1 | string | Conurbation 1 | |
| alignment | string | Alignment | |
| primary_permanent_feature | string | Primary permanent feature | |
| primary_temporary_feature | string | Primary temporary feature | |
| primary_hazardous_feature | string | Primary hazardous feature | |
| street_lighting | string | Street lighting | |
| road_surface | string | Road surface | |
| surface_condition | string | Surface condition | |
| weather | string | Weather | |
| natural_lighting | string | Natural lighting | |
| signals_operation | string | Signals operation | |
| other_traffic_control | string | Other traffic control | |
| speed_limit | string | Speed limit | |
| road_classification_admin | string | Road classification (admin) | |
| rum_code | integer | RUM - code | |
| rum_description | string | RUM - description | |
| dca_code | integer | DCA - code | |
| dca_description | string | DCA - description | |
| dca_supplement | string | DCA supplement | |
| first_impact_type | string | First impact type | |
| key_tu_type | string | Key TU type | |
| other_tu_type | string | Other TU type | |
| traffic_units | integer | No. of traffic units involved | |
| killed | integer | No. killed | |
| seriously_injured | integer | No. seriously injured | |
| moderately_injured | integer | No. moderately injured | |
| minor_other_injured | integer | No. minor-other injured | |
| sa2_2021_code | string | (spine: sa2) | Code of the Statistical Area Level 2 (2021) the point falls in. Joined by location from abs-sa2-2021, not published by the publisher. |
| sa2_2021_name | string | (spine: sa2) | Name of the Statistical Area Level 2 (2021) the point falls in. Joined by location from abs-sa2-2021, not published by the publisher. |
| lga_2025_code | string | (spine: lga) | Code of the Local Government Area (2025) the point falls in. Joined by location from abs-lga-2025, not published by the publisher. |
| lga_2025_name | string | (spine: lga) | Name of the Local Government Area (2025) the point falls in. Joined by location from abs-lga-2025, not published by the publisher. |
| sal_2021_code | string | (spine: suburb) | Code of the Suburb and Locality (2021) the point falls in. Joined by location from abs-suburbs-localities-2021, not published by the publisher. |
| sal_2021_name | string | (spine: suburb) | Name of the Suburb and Locality (2021) the point falls in. Joined by location from abs-suburbs-localities-2021, not published by the publisher. |
| poa_2021_code | string | (spine: postcode) | Code of the Postal Area (2021) the point falls in. Joined by location from abs-postal-areas-2021, not published by the publisher. |
| poa_2021_name | string | (spine: postcode) | Name of the Postal Area (2021) the point falls in. Joined by location from abs-postal-areas-2021, not published by the publisher. |
| sed_2025_code | string | (spine: state_electorate) | Code of the State Electoral Division (2025) the point falls in. Joined by location from abs-state-electoral-divisions-2025, not published by the publisher. |
| sed_2025_name | string | (spine: state_electorate) | Name of the State Electoral Division (2025) the point falls in. Joined by location from abs-state-electoral-divisions-2025, not published by the publisher. |
| ced_2025_code | string | (spine: federal_electorate) | Code of the Commonwealth Electoral Division (2025) the point falls in. Joined by location from abs-federal-electoral-divisions-2025, not published by the publisher. |
| ced_2025_name | string | (spine: federal_electorate) | Name of the Commonwealth Electoral Division (2025) the point falls in. Joined by location from abs-federal-electoral-divisions-2025, not published by the publisher. |
Download this table as a data dictionary (Excel), or read it as a Frictionless Table Schema at schema.json. The key is crash_id.
A sample of 10 rows
From the latest version, newest first by reporting year, each degree of crash in turn, then in the publisher's order, with every field. A blank cell is shown as null.
| crash_id | degree_of_crash | degree_of_crash_detailed | reporting_year | year_of_crash | month_of_crash | day_of_week_of_crash | two_hour_intervals | street_of_crash | street_type | distance | direction | identifying_feature | identifying_feature_type | town | route_no | school_zone_location | school_zone_active | type_of_location | latitude | longitude | lga | urbanisation | conurbation_1 | alignment | primary_permanent_feature | primary_temporary_feature | primary_hazardous_feature | street_lighting | road_surface | surface_condition | weather | natural_lighting | signals_operation | other_traffic_control | speed_limit | road_classification_admin | rum_code | rum_description | dca_code | dca_description | dca_supplement | first_impact_type | key_tu_type | other_tu_type | traffic_units | killed | seriously_injured | moderately_injured | minor_other_injured | sa2_2021_code | sa2_2021_name | lga_2025_code | lga_2025_name | sal_2021_code | sal_2021_name | poa_2021_code | poa_2021_name | sed_2025_code | sed_2025_name | ced_2025_code | ced_2025_name |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1366239 | Fatal | Fatal | 2025 | 2025 | January | Friday | 00:01 - 01:59 | HUME | HWY | 320 | North | SPRING DAM | RD | COOLAC | 2 | No | Not a school zone | Divided road | -34.861642 | 148.285932 | Cootamundra-Gundagai | Country non-urban | Rest of NSW - Rural | Curved | null | Previous crash | null | Nil | Sealed | Dry | Fine | Darkness | Nil | No traffic controls | 110 km/h | State | 62 | Accident | 603 | On path - Previous accident | null | Other angle | B-double (road train prior 98) | Semi-trailer | 4 | 1 | 1 | 2 | 0 | 113031264 | Gundagai | 12160 | Cootamundra-Gundagai | 11022 | Coolac | 2727 | 2727 | 10022 | Cootamundra | 138 | Riverina |
| 1366300 | Injury | Serious Injury | 2025 | 2025 | January | Saturday | 10:00 - 11:59 | CASINO | ST | 0 | Right on the spot | CANIABA | ST | SOUTH LISMORE | 544 | No | Not a school zone | X-intersection | -28.807209 | 153.257206 | Lismore | Country urban | Rest of NSW - Urban | Straight | null | null | null | Off | Sealed | Dry | Overcast | Daylight | Nil | Give way sign | 50 km/h | Regional | 21 | Right through | 202 | Opp - Right-thru | null | Other angle | Car (sedan/hatch) | Motorcycle | 2 | 0 | 2 | 0 | 1 | 112021248 | Lismore | 14850 | Lismore | 13605 | South Lismore | 2480 | 2480 | 10045 | Lismore | 132 | Page |
| 1366366 | Non-casualty (towaway) | Non-casualty (towaway) | 2025 | 2025 | February | Sunday | 04:00 - 05:59 | DUBLIN | ST | 0 | Right on the spot | WINTEN | DR | GLENDENNING | null | No | Not a school zone | T-junction | -33.741797 | 150.859354 | Blacktown | Sydney metro. area | Syd-Newc-Woll Gtr conurbation | Straight | null | null | null | Nil | Sealed | Dry | Fine | Darkness | Nil | No traffic controls | 50 km/h | Local | 81 | Off left/rt bnd=>obj | 707 | Right turn | Left | Other angle | Car (sedan/hatch) | 4 wheel drive / SUV | 2 | 0 | 0 | 0 | 0 | 116031314 | Glendenning - Dean Park | 10750 | Blacktown | 11663 | Glendenning | 2761 | 2761 | 10010 | Blacktown | 108 | Chifley |
| 1366241 | Fatal | Fatal | 2025 | 2025 | January | Monday | 10:00 - 11:59 | SHERBROOK | RD | 100 | South | BURDETT | ST | HORNSBY | 7245 | No | Not a school zone | 2-way undivided | -33.704203 | 151.10605 | Hornsby | Sydney metro. area | Syd-Newc-Woll Gtr conurbation | Straight | null | null | null | Off | Sealed | Dry | Fine | Daylight | Nil | No traffic controls | 50 km/h | Regional | 1 | Ped emerging | 2 | Ped - emerging | null | Vehicle - Pedestrian | Car (sedan/hatch) | Pedestrian | 2 | 1 | 0 | 0 | 0 | 121021579 | Wahroonga (West) - Waitara | 14000 | Hornsby | 11943 | Hornsby | 2077 | 2077 | 10086 | Wahroonga | 104 | Berowra |
| 1366364 | Injury | Moderate Injury | 2025 | 2025 | March | Wednesday | 12:00 - 13:59 | RAMSGATE | RD | 0 | Right on the spot | CAMPBELL | ST | RAMSGATE | 624 | No | Not a school zone | X-intersection | -33.984344 | 151.137481 | Bayside | Sydney metro. area | Syd-Newc-Woll Gtr conurbation | Straight | null | null | null | Off | Sealed | Dry | Fine | Daylight | Nil | Give way sign | 60 km/h | Regional | 11 | Right far | 102 | Adj - Right-thru from left | null | Right angle | Motorcycle | Car (sedan/hatch) | 2 | 0 | 0 | 1 | 0 | 119041382 | Sans Souci - Ramsgate | 10500 | Bayside (NSW) | 13504 | Sans Souci | 2219 | 2219 | 10071 | Rockdale | 109 | Cook |
| 1366374 | Non-casualty (towaway) | Non-casualty (towaway) | 2025 | 2025 | March | Saturday | 06:00 - 07:59 | PACIFIC | HWY | 120 | South | APANIE | AVE | NARARA | 10 | No | Not a school zone | 2-way undivided | -33.392272 | 151.352452 | Central Coast | Country urban | Syd-Newc-Woll Gtr conurbation | Curved | null | null | null | Nil | Sealed | Dry | Fine | Daylight | Nil | No traffic controls | 60 km/h | State | 81 | Off left/rt bnd=>obj | 803 | Off right bend into obj | Left | Other angle | Car (sedan/hatch) | 4 wheel drive / SUV | 2 | 0 | 0 | 0 | 0 | 102011043 | Wyoming | 11650 | Central Coast (NSW) | 12905 | Narara | 2250 | 2250 | 10081 | The Entrance | 112 | Dobell |
| 1366245 | Fatal | Fatal | 2025 | 2025 | January | Monday | 10:00 - 11:59 | PUNCHBOWL | RD | 10 | East | ELLIOTT | ST | BELFIELD | 549 | No | Not a school zone | T-junction | -33.900702 | 151.086036 | Strathfield | Sydney metro. area | Syd-Newc-Woll Gtr conurbation | Straight | null | null | null | Off | Sealed | Dry | Fine | Daylight | Nil | No traffic controls | 60 km/h | State | 71 | Off rd left => obj | 703 | Left off cway into object | null | Vehicle - Object | Car (sedan/hatch) | null | 2 | 1 | 2 | 0 | 0 | 119021362 | Belmore - Belfield | 11570 | Canterbury-Bankstown | 10257 | Belfield | 2191 | 2191 | 10075 | Strathfield | 143 | Watson |
| 1366367 | Injury | Serious Injury | 2025 | 2025 | March | Thursday | 10:00 - 11:59 | OLYMPIC | HWY | 70 | North | WALLENDOON | ST | COOTAMUNDRA | 78 | No | Not a school zone | 2-way undivided | -34.641323 | 148.029209 | Cootamundra-Gundagai | Country urban | Rest of NSW - Urban | Straight | B'down lane/road shldr | null | null | Off | Sealed | Dry | Fine | Daylight | Nil | No traffic controls | 50 km/h | State | 42 | Leaving parking | 401 | Manov - Leaving parking | null | Other angle | 4 wheel drive / SUV | 4 wheel drive / SUV | 2 | 0 | 1 | 1 | 0 | 113031263 | Cootamundra | 12160 | Cootamundra-Gundagai | 11053 | Cootamundra | 2590 | 2590 | 10022 | Cootamundra | 138 | Riverina |
| 1366381 | Non-casualty (towaway) | Non-casualty (towaway) | 2025 | 2025 | March | Tuesday | 10:00 - 11:59 | CARTER | ST | 0 | Right on the spot | ROKETTO BADMINTON | ENT | LIDCOMBE | null | No | Not a school zone | 2-way undivided | -33.850443 | 151.058034 | Parramatta | Sydney metro. area | Syd-Newc-Woll Gtr conurbation | Straight | Driveway/entrance | null | null | Off | Sealed | Dry | Fine | Daylight | Nil | No traffic controls | 60 km/h | Local | 32 | Right rear | 303 | Same - Rear right | null | Rear end | 4 wheel drive / SUV | Car (sedan/hatch) | 2 | 0 | 0 | 0 | 0 | 125011710 | Wentworth Point - Sydney Olympic Park | 16260 | Parramatta | 12325 | Lidcombe | 2141 | 2141 | 10064 | Parramatta | 136 | Reid |
| 1366247 | Fatal | Fatal | 2025 | 2025 | January | Tuesday | 14:00 - 15:59 | MARTINDALE | RD | 0 | Right on the spot | SMITHS | BDGE | MARTINDALE | null | No | Not a school zone | 2-way undivided | -32.464394 | 150.667477 | Muswellbrook | Country non-urban | Rest of NSW - Rural | Curved | Narrow/1-lane bridge | null | null | Nil | Sealed | Dry | Overcast | Daylight | Nil | Other traf. control | 100 km/h | Local | 20 | Head on | 201 | Opp - Head on | null | Head-on | Large rigid | 4 wheel drive / SUV | 2 | 1 | 1 | 1 | 0 | 106041127 | Muswellbrook Surrounds | 15650 | Muswellbrook | 12520 | Martindale | 2328 | 2328 | 10083 | Upper Hunter | 130 | New England |
The first row as JSON
As it appears in data.ndjson.
GET https://publicdata.au/d/nsw-road-crashes/latest/data.ndjson
{
"crash_id": 1244876,
"degree_of_crash": "Injury",
"degree_of_crash_detailed": "Serious Injury",
"reporting_year": 2021,
"year_of_crash": 2020,
"month_of_crash": "October",
"day_of_week_of_crash": "Friday",
"two_hour_intervals": "12:00 - 13:59",
"street_of_crash": "GREGORY HILLS",
"street_type": "DR",
"distance": 0,
"direction": "Right on the spot",
"identifying_feature": "CAMDEN VALLEY WAY",
"identifying_feature_type": "TO",
"town": "GREGORY HILLS",
"route_no": null,
"school_zone_location": "No",
"school_zone_active": "Not a school zone",
"type_of_location": "Divided road",
"latitude": -34.020436,
"longitude": 150.76167,
"lga": "Camden",
"urbanisation": "Sydney metro. area",
"conurbation_1": "Syd-Newc-Woll Gtr conurbation",
"alignment": "Straight",
"primary_permanent_feature": null,
"primary_temporary_feature": null,
"primary_hazardous_feature": null,
"street_lighting": "Off",
"road_surface": "Sealed",
"surface_condition": "Dry",
"weather": "Fine",
"natural_lighting": "Daylight",
"signals_operation": "Nil",
"other_traffic_control": "Ped crossing",
"speed_limit": "60 km/h",
"road_classification_admin": "Local",
"rum_code": 34,
"rum_description": "Lane change right",
"dca_code": 306,
"dca_description": "Same - Lane change right",
"dca_supplement": null,
"first_impact_type": "Other angle",
"key_tu_type": "Pedal cycle",
"other_tu_type": "Station wagon",
"traffic_units": 2,
"killed": 0,
"seriously_injured": 1,
"moderately_injured": 0,
"minor_other_injured": 0,
"sa2_2021_code": "127011727",
"sa2_2021_name": "Gledswood Hills - Gregory Hills",
"lga_2025_code": "11450",
"lga_2025_name": "Camden",
"sal_2021_code": "11780",
"sal_2021_name": "Gregory Hills",
"poa_2021_code": "2557",
"poa_2021_name": "2557",
"sed_2025_code": "10013",
"sed_2025_name": "Camden",
"ced_2025_code": "125",
"ced_2025_name": "Macarthur"
}
Query it without downloading
The query API returns only the rows you ask for, as JSON, NDJSON or CSV, from the same typed rows as the files. It needs no key. Build a query here and run it, then copy the URL or the code into your own work.
How filters work
Each filter is field=operator.value in the query string, and every filter must match. A number or date field compares as a number or date, and a text field compares as text. Values are case-sensitive, except that like and ilike ignore case in the ASCII letters A to Z. Prefix not. to negate a filter, as not.eq.value.
| Operator | What it matches |
|---|---|
| eq.value | Equal to the value. |
| neq.value | Not equal to the value. A blank cell does not match. |
| gt.value | Greater than the value. |
| gte.value | Greater than or equal to the value. |
| lt.value | Less than the value. |
| lte.value | Less than or equal to the value. |
| like.*text* | Matches a pattern where * stands for any run of characters, ignoring case in the ASCII letters A to Z. |
| ilike.*text* | The same as like. Both ignore case in ASCII letters only, so É and é differ. |
| in.(a,b,c) | Equal to any value in the list. A value cannot contain a comma. |
| is.null | Blank in the source, or suppressed by the publisher. |
| Parameter | What it does |
|---|---|
| select | Fields to return, comma-separated. Every field when absent. |
| order | field.asc or field.desc, comma-separated. When absent, rows come in the order of the file they are read from. For the newest versions that is the version's own data.parquet, which keeps the publisher's order unless the dataset sorted it. Any other version is read from the Parquet file the answer names in file, and when that file is sorted the answer names its sort in order, which ties follow too. |
| limit | Rows per page, 1 to 10,000. 100 when absent. |
| offset | Rows to skip. The next URL in each answer sets it for you. |
| group | On aggregate, fields to group by, comma-separated. |
| metric | On aggregate, count, sum.field, avg.field, min.field or max.field, comma-separated. count when absent. |
| format | json, ndjson or csv. JSON carries the provenance header, and the others carry it in response headers. |
Without a version the API answers from the newest version, and that answer changes when the publisher releases again. Put versions/<date>/ before rows or aggregate for an answer from that version alone. Every version with a table can be queried, and /api/v1/datasets/<slug>/versions lists them. A query on a version the database does not hold is read from its Parquet file, and one that would read too much of it is refused with DuckDB SQL that answers it from the file. The 2 newest versions are loaded; versions lists them.
The API allows 60 requests in 10 seconds from one address. Above that it answers 429 for 10 seconds with a Retry-After header, a RateLimit-Policy header and a JSON body that gives the limit. Every API answer carries the same RateLimit-Policy. A client should wait the Retry-After seconds, or read the files, which have no limit. openapi.json describes this dataset's query paths for client generators and agents.
https://publicdata.au/mcp is a remote MCP server over Streamable HTTP with the same tools, defined in the same place, so a tool added to the pages is added here too. It needs no key and no account. Its row tools and the query API answer every version through one query path, so a tool's answer and the query URL it cites give the same rows. A query that would read too much of a file is refused with DuckDB SQL that answers it from the file. It also has search and fetch, the two tools ChatGPT's deep research and company knowledge call by name, which run search_datasets and read a dataset's record in the shapes those features expect. The row tools are held to the query API's limit of 60 queries in 10 seconds from one address, and each call is two queries because it also counts the matching rows. Add it to Claude Code with claude mcp add --transport http publicdata https://publicdata.au/mcp, add the URL in Claude as a custom connector, connect the publicdata.au plugin in ChatGPT, or give it to any client that connects to remote MCP servers. Each dataset with a table is also a resource at https://publicdata.au/d/<slug>/fields.json, which lists its fields with their types, the publisher's descriptions, their ranges and the values they hold, so an agent can read a dataset's shape before it writes a query.
Smaller files
The same rows split by a field, one JSON file each with a GeoJSON beside it. Each file carries the same provenance header.
By year_of_crash (6 files)
by/year_of_crash/index.json lists every file with its row count.
What changed
One version for every release the publisher has made since this site started following the dataset. The date is the day the file changed on the portal.
- 2026-10-09 93,042 rows+19557 added −18704 removed 139 changed62 fieldsxlsx7bbceb2f94cd
- 2025-10-10 92,189 rows62 fieldsxlsx7345189f017d
versions.json · changes.json · history.tar.zst (8.3 MB, every version's Parquet and manifest)
Questions
How do I download Road crashes as a CSV file?
Open https://publicdata.au/d/nsw-road-crashes/latest/data.csv. It redirects to the newest dated version, which is https://publicdata.au/d/nsw-road-crashes/v/2026-10-09/data.csv today. The same path serves Parquet, SQLite, DuckDB, GeoPackage, GeoParquet, NDJSON. A dated URL never changes, so use it when the file must stay the same.
Can I open Road crashes in Excel?
Not as a workbook. Excel is not offered because the table is 58.1 MB as CSV, over the 50 MB limit for Excel. Load the CSV or the Parquet file with Power Query, or take one partition file at a time. https://publicdata.au/d/nsw-road-crashes/v/2026-10-09/data.csv.gz is the CSV at about a tenth of the size.
What years does Road crashes cover?
The current version covers 2021 to 2025. Each release from TfNSW becomes a new dated version here, and earlier versions stay online.
How often is Road crashes updated?
TfNSW releases it yearly. This site checks the portal every week and adds a dated version when the file changes.
Can I use Road crashes commercially?
Yes. CC BY 3.0 AU allows commercial use, redistribution and derived works as long as the attribution is kept. The attribution string is in this page's side column and inside every file.
Is this the official source for Road crashes?
No. The publisher is Transport for NSW, and its page is https://data.nsw.gov.au/data/dataset/2-nsw-crash-data. National Digital converted the publisher's file to other formats. It renamed and typed the columns and kept only those the schema lists. It added place columns from ABS boundary layers, which the schema marks as joined. National Digital made these changes, and the publisher is not responsible for them. The original sits beside every version as source.xlsx with its SHA-256, so the two can be compared.
How do I get only the rows for one year of crash?
Every version has one JSON file per value of year_of_crash, 6 files in the current version, listed with row counts at https://publicdata.au/d/nsw-road-crashes/v/2026-10-09/by/year_of_crash/index.json. For example https://publicdata.au/d/nsw-road-crashes/v/2026-10-09/by/year_of_crash/2025.json holds the 19,506 rows where year_of_crash is 2025. A GeoJSON file sits beside each one.
What coordinate system does Road crashes use?
Latitude and longitude in decimal degrees as published. The publisher's data manual does not state the datum; GDA94 is the NSW Government's standard and differs from GDA2020 and WGS84 by under 2 metres. The GeoJSON file uses the same coordinates, which is what GeoJSON expects.
Which SA2, council area, suburb, postcode, state electorate and federal electorate is each row of Road crashes in?
Each row with coordinates carries the SA2 (2021), Council area (LGA 2025), Suburb or locality (2021), Postcode (postal area 2021), State electorate (2025) and Federal electorate (2025) its point falls in, with the code of each. This site joins them by location against the ABS boundaries, https://publicdata.au/d/abs-sa2-2021/, https://publicdata.au/d/abs-lga-2025/, https://publicdata.au/d/abs-suburbs-localities-2021/, https://publicdata.au/d/abs-postal-areas-2021/, https://publicdata.au/d/abs-state-electoral-divisions-2025/, https://publicdata.au/d/abs-federal-electoral-divisions-2025/, and TfNSW did not publish them. The schema marks each one as joined and names the boundary version. A row without coordinates, or a point outside every area, has them blank. Place columns from Australian Statistical Geography Standard (ASGS) Edition 3 boundaries, Australian Bureau of Statistics, licensed under CC BY 4.0.
What this site did to the data. Cells were typed, headers were renamed and the encoding was made UTF-8. Rows were left alone. The publisher's file sits beside every version as source.xlsx so the change can be checked.
The licence requires the publisher's attribution. We ask that you also name publicdata.au and link to the version you used.
Transport for NSW, NSW Government, NSW Crash Data, sourced 12 October 2026, https://data.nsw.gov.au/data/dataset/2-nsw-crash-data, licensed under CC BY 3.0 AU. Serialised and versioned by National Digital at publicdata.au, https://publicdata.au/d/nsw-road-crashes/v/2026-10-09/
The map's state outlines are from the ABS Australian Statistical Geography Standard 2021, under CC BY 4.0.
The attribution alone is inside every file. Harvard is the author-date form a report or a brief asks for. Paste the HTML on a web page so the links carry.
Transport for NSW states the licence, CC BY 3.0 AU, on its own page, which was read and recorded here on 29 September 2026.
Place columns from Australian Statistical Geography Standard (ASGS) Edition 3 boundaries, Australian Bureau of Statistics, licensed under CC BY 4.0.
Every fetch reads the licence again and stops when it has changed. The fetch that made version 2026-10-09 read it from data.nsw.gov.au on 12 October 2026.
- Hugging Faceload_dataset("National-Digital/nsw-road-crashes")
- Kagglewith a starter notebook
- ZenodoDOI 10.5281/zenodo.23077195
Each copy carries the same attribution and links back to this page.
- datapackage.jsonFrictionless data package pointing at the latest version
- schema.jsonTable Schema for every field
- schema.xlsxthe same field list as an Excel data dictionary
- versions.jsonevery version with date, rows, fields and hash
- changes.jsondiff for every consecutive pair
- manifest.jsonwhere the bytes came from and when
- Query APIfiltered rows and counts without a download
- openapi.jsonOpenAPI 3.1 for this dataset's query paths
- explore/charts and dashboards in the browser, shareable and embeddable
- index.mdthis page as Markdown
This page carries a schema.org Dataset record and is listed in /.well-known/ard.json and /catalog.json.
Transport for NSW publishes this table on opendata.transport.nsw.gov.au. Questions about the content go to the publisher. Questions about the files here go to corrections.
Fetched 12 October 2026. Source SHA-256 7bbceb2f94cd9f36…
This page is built from register/nsw-road-crashes.yaml on GitHub. Anyone can propose a fix to it, such as a clearer description or a field the publisher renamed, as a pull request.
Report a file that differs from the publisher's · How to contribute