A suite of software tools is presented for managing a large parallel programming project. The tools were selected recognizing that parallel program development is an iterative process and subject to mistakes and that software tools can be useful for maintaining source code flexibility and portability, tracking revisions, and analyzing variable usage and loop structure within a program. The tools discussed are: make, cpp, RCS, and FORGE 90. The concept of a toy program is introduced as a means for experimenting with a simpler version of an application program. Finally, the use of these tools and techniques is demonstrated as part of an optimization and parallelization effort for a scientific application program called ZELIG.