Having just had to import data that included a membership number for a client, rather than make a new field for the number it made sense to use the Contact ID, given that this is already a consecutive number and unique to each contact. The problem was that over the years there were big gaps in the consecutive numbers.
