>The first thing I thought of when I read your original message was SET EXACT ON/OFF. However, based on what you posted, I'm not sure that would come into play (cuz you're comparing SUBSTR(...,1,1)="C" etc).
>
>But there must be some SET setting different between your development environment and your EXE environment. Can you do a LIST STATUS before the DELETE command in both instances and see if there's any difference?
>
>The other thing that comes to mind is an old esoteric phenomenon where a SUBSTR function acting beyond the end of the string will not produce an error message in development, but it *will* in runtime (or else it had something to do with the setting of TALK or some such... I can't remember). I'm looking at your SUBSTR(ALLTRIM()) here. If there's an empty string, then looking at SUBSTR(...,1,1) will be technically illegal.
>
>Could this be coming into play somehow?
>
>For what it's worth, the way I'd code it would be like so:
>
>SET EXACT OFF
>DELETE FOR LTRIM(jtspart)="C" OR jtspart0="=" or "|"$jtspart0 or ","$LEFT(jtspart0,29)
>
>--Brad

Thanx Brad,

After a night of sleep, I found a logic flaw.
Final Data is processed from records that DO contain the pipe and do-not start with C. I looped the potentials and processed only the needed. No Delete needed therefore.

The VFP EXE now performs effeciently from any call except from MSACCESS.
Another thread for that problem : error - cannot find .

I like that List Status, Thank you both for your help
Ed B

Edgar L. Bolton, B.S. B.B.A.