I need help producing a query which will give me unique city values based on the state.

Sample Structure.

id|state|city|zip

1|CO|Denver|77201
2|CO|Denver|77202
3|CO|Denver|77203
4|IL|Chicago|60505
5|IL|Chicago|60504
6|IL|Springfield|61202
7|IL|Joliet|65243
8|CO|Arvada|77302
9|IN|Springfield|50309
10|IN|Hamilton|50398
11|DE|Denver|10348


Desired results:
CO Denver
IL Chicago
IL Springfield
IL Joliet
CO Arvada
IN Springfield
IN Hamilton
DE Denver

What I have tried so far.
SELECT DISTINCT(city) FROM states.
That gives me:
CO|Denver|77201
IL|Chicago|60505
IL|Springfield|61202
IL|Joliet|65243
CO|Arvada|77302
IN|Hamilton|50398

As you can see it removed DE's denver, I only want it to remove duplicate cities for each state not duplicate cities for the entire US.

Also tried using GROUP BY city same result.
Also tried GROUP BY state and that only gives one result per state.

Using php I can just pull all the results and and remove the duplicates. This gives me the desired results but my database has nearly 50000 records so it takes forever.
If there is a simple query I can do that would be great. Any help or ideas would be greatly appreciated.

I used a table with the same sample data and was able to get

SELECT DISTINCT(state), city FROM states

to work.

Frankly I would use two tables to store this anyway, to reduce the storage size.

id|state|city

1 |CO   |Denver
4 |IL   |Chicago
6 |IL   |Springfield
7 |IL   |Joliet
8 |CO   |Arvada
9 |IN   |Springfield
10|IN   |Hamilton
11|DE   |Denver

id|zip

1 |77201
1 |77202
1 |77203
4 |60505
4 |60504
6 |61202
7 |65243
8 |77302
9 |50309
10|50398
11|10348

Now not only is your query easier, but you don't store as much duplicate data.
Now to add a new zip code, just add the id and zip to the second table, rather than the id,state,city, and zip to the first table.