speeding up a join COUNT

[James Tu]

I'm performance testing my 'users' table.  It currently has roughly  
1M user records.  The 'geo_entities' table has ~ 250 records.

Here's my query.

SELECT users.entity_id, geo_entities.entity_name,  
geo_entities.short_code, COUNT( users.entity_id )
FROM users, geo_entities
WHERE users.user_type = 'user'
AND users.entity_id = geo_entities.id
GROUP BY entity_id
LIMIT 0 , 30

It took 51 seconds to execute.

Both tables only have an index on their unique record id.
Is there a way to speed up this up?

-James