Hi,
I have a simple table like CREATE TABLE RecipeIngredients ( id integer primary key, recipeid integer(10), unitId integer(10), ... ); With an index on recipeid and unitid (foreign keys) In the table there are 11 rows where recipeid is "558134" and 243042 rows where the unitid is "3" from totally 5.4 mill. rows. If I do a select recipeid, unitid from RecipeIngredients where recipeid = 558134 and unitid = 3 it takes about 80 seconds. If I change the order in the expression to select recipeid, unitid from RecipeIngredients where unitid = 3 and recipeid = 558134 it also takes about 80 seconds. But if I do select recipeid, unitid from RecipeIngredients where unitid > 3 and recipeid = 558134 It takes less than 1 second. I tried to do a "explain" and can see that there are a difference in the way it is looping through data, but don't really understand why. Any one has a good explanation? Frank Tolstrup