Have you tried process of elimination by "brute-force"? You could take a handful of the PRGs that might be suspect (start with the largest ones) and replace with a "stub" with only header part and re-compile and see if you see any difference. Another thing you could watch out for are include files. I'd also run into on occasion where something went awry in code generation of .MPR files from menus that caused similar problems. Back in the FPD/FPW days, I'd seen on occasion very complex screen files that would result in .SPR files that would be too large to compile -- especially if they utilized include files extensively.

>Attempting to compile the latest change and build a new .EXE, I received Error 1202: Program is too large. Help says there is a 65K limit on single procedure or program, but provides no way to identify which file.
>
>The previous build of the executable did not throw the error, despite the .EXE being 23 MB. and ADIR() showing half a dozen .PRGs and two .FXPs over 65K.
>
>The most likely culprit was split into two files, but the error persists. VFP 9 is available in development but cannot be used for deployment yet. I'm attempting to use this as a push to move to VFP 9, but for now, I need a VFP 6 solution to the problem.
>
>Chris.