Do you remember how many rows the source table had prior to doing the deletions and either what the size of the Clustered Index was or what the average row length was? I’d like to do a little testing with an alternate method. It might not amount to a thing but I won’t know until I test it.

Thanks, and thanks for the article. The “middle ground” (as you described it) is a tried an true method of doing such large deletes.