I have succesfully impimented the RM1 method in 
  Gaussian 03. All values which need to be changed
  can be found under the heading :

    START OF AM1 PARAMETERS
  
   in the utilam.F file in the g03
  directory.

 I have attached a spread sheet of all of the values
 which need to be changed:
 RM1_Parameters.xls

 
 My preliminary results with the method show 
 significant improvements over AM1 methods, in
 particular for guanidinium systems, and in general
 pi-systems involving N atoms.


 Keith T. Butler