Skip to content
Geo Hound
Add to Chrome
Guide

How to find the WMS URL behind a web map

The short answer: the map already fetched it, so the URL is in your browser's network traffic. Open DevTools, filter the requests for GetMap or WMS, and the service URL is the part before the ?. Everything below is the detail, plus two faster ways.

A web map viewer almost never tells you where its data comes from. There is no "copy service URL" button, because the people who built the viewer were building a viewer, not a data portal. But the map cannot draw itself without asking a server for pixels, and that request is visible to you.

Method 1: the network tab

This is the method everyone learns first, and it always works.

  1. Open the web map and let it finish loading.
  2. Press F12 to open developer tools, and click the Network tab.
  3. Reload the page. You need the requests that happen during load, and the panel only records while it is open.
  4. In the filter box, type GetMap. A WMS layer draws itself by calling GetMap, so this cuts hundreds of requests down to a handful.
  5. Click one and look at its URL.

You are looking at something like this:

https://maps.example.govt.nz/services/wms?SERVICE=WMS&VERSION=1.3.0
  &REQUEST=GetMap&LAYERS=flood_zones&CRS=EPSG:3857
  &BBOX=19430000,-5000000,19440000,-4990000&WIDTH=256&HEIGHT=256&FORMAT=image/png

The service URL is everything before the ?. In this case: https://maps.example.govt.nz/services/wms. The rest is just this particular request: which layer, which area, how big. Throw it away.

The other thing worth noting is LAYERS=flood_zones. That is the layer name you will need later, and it is frequently different from the friendly label the viewer showed you.

If GetMap finds nothing, the map is probably not WMS. Try filtering for WMTS, FeatureServer, MapServer, or .pbf. See how to get data out of an ArcGIS web map and how to find the XYZ tile URL.

Method 2: ask the service what it has

Once you have the service URL, ask it to describe itself:

https://maps.example.govt.nz/services/wms?SERVICE=WMS&REQUEST=GetCapabilities&VERSION=1.3.0

You get an XML document listing every layer the service publishes, the projections it supports, the bounding boxes, and the image formats. This is the difference between having one layer and having the whole service.

It is worth knowing that this often reveals layers the public viewer never displayed. A council viewer might show you three layers while the service publishes forty. That is not a security hole: the service publishes them deliberately. It just means the viewer was curated.

There is a guide to reading a GetCapabilities document if the XML looks impenetrable.

Method 3: let something watch for you

The network tab works, but it is tedious, and it is not something you want to do on every map you look at. That is what Geo Hound does: it watches the traffic while you browse, recognises WMS, WFS, WMTS, WCS, OGC API, and the Esri REST family, and saves each one with its title, provider, layer list, coordinate systems, and bounding box already fetched from the capabilities document.

Then it hands you the URL already formatted for QGIS, so you skip the cleanup step entirely.

Getting it into QGIS

With the service URL in hand:

  1. In QGIS, open the Browser panel.
  2. Right-click WMS/WMTS and choose New Connection.
  3. Paste the service URL, the bit before the ?. Give it a name.
  4. Click OK, then expand the connection to see the layers.

If QGIS says it cannot connect, the usual suspects are: you pasted the whole GetMap URL rather than trimming it, the service needs a version parameter, or the service is behind a login your browser had and QGIS does not.

A note on being a good citizen

Finding a service URL is not hacking. You are reading a request your own browser made to a public server, and that server answered because it was configured to answer anyone.

That said, being able to fetch the data is not the same as being allowed to use it. Public map services usually carry a licence, and the terms are typically on the portal that hosts the viewer. Check before you republish, and cache what you need rather than hammering someone's server with automated requests. These are frequently small organisations paying for their own bandwidth.