>I didn't extract the code to do a separate test, but I ran the coverage profiler on the program in VFP and the line that take the most time is the Word.Print command.
>
>>Did you compare the manual process with a automated program to just do the mailmerge (e.g. starting from creating word object) ?

The reason I want to do it as a separate program is to eliminate anything else that may already affected the memory. E.g. I want to run a clean test of manual MailMerge vs. MailMerge through the VFP.