Suppose I have very big table and I need to run query:

select * from table where field1 = @value1 and field2 = @value2

Which index(es) would be better for the table?
Separate indexes for both field1 and field2 or one (maybe clustered) index with field1+field2?
Which one? and why?