Software maintenance is the process of designing and integrating consistent changes to an existing software system. Doing effective maintenance is a complex task due in part to the size and complexity of the maintained software, and the widely varying maintenance support tools and environments that are available. These problems are compounded since most existing software produced prior to the use of structured techniques. It is difficult for the maintainer to ascertain the complete effect of a code change. Decomposition slicing yields guidelines for maintainers to use so that changes can be assured to be completely contained in the components under consideration and that there are no undetected linkages between the modified and unmodified code. Thus the impact of changes can be assessed. The decomposition can also be used to limit the amount of testing required to assure that the change is correct. The goal of this research is to scale up this methodology and validate its effectiveness on large software systems by an empirical evaluation.