The bond strength of polymer interfaces within packaged microelectronic devices significantly influences their reliability. In the interest of predictive modeling and to facilitate materials selection during the design process, it is highly desirable to be able to distinguish between the adhesive performances of multiple polymer interfaces. However, typical adhesion testing is normally plagued by large deviations in its test results which make drawing statistical conclusions from adhesion strength data difficult. To remedy this, an investigation into the primary sources of variance associated with the pull test was performed. Four primary factors were identified, load alignment, loading rate, bond thickness, and the edge condition. The control of each of these four parameters was targeted during the development of an improved adhesion test technique. The results are an adhesion measurement method which has successfully reduced the scatter in test results from a standard deviation of 50% to approximately 10%.