This seems to compile and run just fine:

* Test code *
x = 1
? x
ENDPROC

...although I think it's sort of a bad practice to do that.

>Is it possible to have a program in which there is an ENDPROC, but no beginning PROC statement? In the program which I decompiled this morning, using ReFox XII (yes, a shameless Plug here! Thanks Jan Brebera!), I noticed that there are several programs with no PROC statement at the beginning of each of these program files. Could it be that they never existed or that the reverse engineering process cut off the PROC statements?
>
>Is this kosher? I mean, will a program file work without a beginning PROC statement if it has an ending ENDPROC statement? I know that a prg file does not need the PROC and ENDPROC unless you write several procedure files inside the prg file.