I want to create a map with openlayers, that contains a vector layer, with all "counties" of all countries of the world. "County" doesn't mean any strict political term here. I just want to split the earth into areas of ~500 - ~2500 km2 (or even much larger in sparsely populated areas), while these areas should have names, that are common to the people living there.

Now I'm searching data sources.

  • The data should be free
  • Each county information should contain the name in one of the languages spoken in this county.
  • It is not important, that the borders are exact - it is just used for a visual selection of some counties, the user wants to do something with
  • It must be complete for each country, that is available (no areas without a "county").
  • For now, it would fit to have those data available for Germany, Austria and Switzerland, but it would be fine to find a solution, that would allow us, to add other countries without having another data import and maintainance workflow for each country.
  • Please let me know, why you downvoted my question, so I can improve it.
    – stofl
    Commented Mar 4, 2014 at 10:20

5 Answers 5


For Germany, Austria and Switzerland and other European countries, you can find the data from the Eurostat (different levels of NUTS). Those are the "official ones". You have 1 294 entities at level 3.

  • Thank you. GADM is more detailed and will be better for my needs, I think.
    – stofl
    Commented Mar 10, 2014 at 15:26

http://www.gadm.org/home has one for whole world in vector format. You download data in several formats. Shp file format may have some problems with column length and local names. But if you use english names that doesnt show up

  • I've seen that website. But I've read "Administrative areas in this database are countries and lower level subdivisions such as provinces, departments, ...". I think, those "counties" are much too large for my needs. I've edited my question to better define, what I mean by "county"
    – stofl
    Commented Mar 3, 2014 at 12:12
  • 1
    I think that provinces and departments should fit to most of your purposes. Of course their size will vary depending on the area of the country, their population density and many other parameters. for instance, in Europe we talk about NUTS to describe the administrative levels. What NUTS level are you looking for ?
    – radouxju
    Commented Mar 3, 2014 at 13:31
  • LAU-1 or LAU-2. If I cannot find something in these levels, NUTS-3 would be okay ... I think. But that would be suboptimal.
    – stofl
    Commented Mar 3, 2014 at 17:04
  • I think, I'll use GADM. Thank you! I thought, that the areas would be larger, as I read the text on the homepage.
    – stofl
    Commented Mar 10, 2014 at 13:41

The GADM boundaries are pretty good, however you'd probably want to review them on a country-by-country basis.

GADM is agnostic about what the administrative units are called, just referring to them as administrative levels one through N. If your goal is to get roughly 2,500 km^2 units, just divide each country's total area by counts at the different levels to see which level comes closest to your target size. You could do that in a simple spreadsheet.

  • Thank you! Do you have a tip, to create that spreadsheet fast?
    – stofl
    Commented Mar 10, 2014 at 13:42
  • In Wikipedia see the article "Table of administrative divisions by country". Go into Wikipedia's article edit mode and copy/paste the underlying HTML-like language into a plain text file on your 'puter. Then edit out the stuff you don't want in your spreadsheet.
    – LADave
    Commented Mar 12, 2014 at 6:27
  • Ah, okay thank you. I think I'll be faster, when I load the shapefile data into a database and synchronize that with a database table created from a wikipedia table, that contains the arears of each country. I thought, you would have a trick, to speed this up.
    – stofl
    Commented Mar 12, 2014 at 17:05
  • I could load the shapefile into a MySQL database with ogr2ogr now. The query to get the table of number of shapes of each area is: select name_0, count(distinct name_1), count(distinct name_2), count(distinct name_3), count(distinct name_4), count(distinct name_5) from features group by name_0;
    – stofl
    Commented Mar 12, 2014 at 17:54
  • Then I could generate a MySQL query from this page to get a list with iso 3166 codes, population and area of each country. This query shows, that almost every data is there now: select distinct features.iso, countries.* from features left join countries on features.iso = countries.isoAlpha3
    – stofl
    Commented Mar 12, 2014 at 18:21

You best bet is: http://www.naturalearthdata.com/

Has administrative area shapefiles.

  • Than you. I've seen this website. But I cannot find data, that fits my requests there. What download do you mean?
    – stofl
    Commented Mar 3, 2014 at 12:17
  • Did you see this data set? naturalearthdata.com/downloads/10m-cultural-vectors/…
    – Chris P
    Commented Mar 3, 2014 at 12:51
  • Yes, but the title says "States, Provinces". I think, that's not, what I'm looking for. Maybe in some countries (Switzerland for example) states or provinces are very small areas. But I think in most (?) countries, they are a lot larger. But I'll try to find a way to preview that shapefiles anyhow.
    – stofl
    Commented Mar 3, 2014 at 13:18
  • I've looked into that data now and the areas are very much too large (for example Bavaria with 70.000km2).
    – stofl
    Commented Mar 10, 2014 at 15:30

I can give you 3 more options:

1.- OpenStreetMap. You can export the administrative boundaries with some tools.

2.- Quattroshapes

3.- GeoPlanet (Yahoo)

You can download it from these link: http://ydn.zenfs.com/site/geo/geoplanet_data_7.9.0.zip

Here you can check the info: http://isithackday.com/geoplanet-explorer/

And, of course, these 2 you already know:

4.- GADM.

5.- Natural Earth

I hope you find them helpful...

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.