I seem to remember reading that COUNT() is faster than COUNT(column), because COUNT(column) worries about NULLs in the column whereas doesn't.
If the column is indexed (it's an auto increment so it should be!) and rows are never deleted, SELECT MAX(id) FROM mytable might be faster still!
MySQL's strength is it's speed, so it probably wont make much difference until you have millions of rows anyway.
Cheers,
Alan