>Based on that, you *should be* ok to make it a clustered index. One thing I haven't seen mentioned here. A clustered index is really a physical storing of the data in index order. You can insert, but it could mean rewriting a large portion of data.
>
>>The ID is an Identity column. If a user creates an order it is (ID) automatically set to the next ID sequential number. If user does not save the newly created record, the "missed" number will never be inserted.
Exactly - if the definition of the clustered index represents a level of randomness when inserting new rows, you potentially wind up with index fragmentation, which can really impact performance.