The research addresses the issue of productivity in application software maintenance. Specifically, it examines the effect of diversity in tools, techniques, hardware and software associated with the portfolio being maintained. In manufacturing environments, there is some evidence to suggest that production of products where there is little sharing of inputs and production processes reduces focus and results in lower manufacturing performance (Skinner, 1974). In economics, it is argued that there are cost complementarities or economies of scope in sharing common inputs and processes among various products with commonalities in production, and diseconomies of scope when inputs and processes differ (Panzar and Willig, 1977, 1981). In the software maintenance context, the issue of diversity and its effect on productivity is particularly salient. Software maintenance is work done to enhance software functionality, correct errors and improve the performance of software (Schneidewind, 1987). It is a costly activity for organizations, requiring from 50 to 80% of the Information Systems (IS) budget and representing more than threefourths of software costs on a life cycle basis (Arthur, 1988). Application portfolio diversity, i.e.,differences in technical platforms, software languages, and development tools and techniques in the set of the organization's software systems, arises as a consequence of the organization's information technology infrastructure decisions. To meet a particular customer need, an IS group acquires or develops software using a certain tool, methodology, and hardware platform. However, it may be that the software does not fit well into the organization's existing technical platform. Furthermore, the software may have been developed using a different methodology or tools than other software systems in the organization's portfolio. This diversity may have the result of increased difficulty in software maintenance because software enhancement can require modification of multiple software systems that have been created using a variety of languages, tools and techniques. The results of our analysis suggest that software portfolio diversity reduces productivity in software maintenance. Potential inefficiencies from diversity in software maintenance can arise from several causes. Switching costs are incurred due to multiple, varied process flows and frequent change over in processes required when modifying software created using different methodologies and tools. Diversity may also increase the difficulty of software quality control, testing and verification; for example, inefficiencies may occur due to the complexities of conducting system and integration testing across multiple technical platforms. Finally, there may be costs due to the difficulties in selecting project team members with the multiple and varied skills required to modify diverse sets of software.