I am currently running one script between two databases. This means that halfway through the script, I need to run mysql_select_db() and then farther down in the code, run it again to switch back to the original database. My question, essentially, is if the overhead in doing this is hideously bad? How many resources could be saved in keeping all of my tables in one database at the cost of disorganization and a database with 200+ tables, many of which have no relationship to each other. It's always been my understanding that a relational database should be just that - tables that all have relationships with each other.