Health and Usage Monitoring Systems (HUMS) is an emerging technology in helicopter aviation. The United States Navy is evaluating its viability for use on its helicopter fleet. HUMS uses sensors placed throughout the helicopter to monitor and record vibration signals and numerous other aircraft operating parameters. This thesis evaluates the vibration signals recorded by a HUMS system using a statistical technique called tree structured classification. The goal of the analysis is to demonstrate the technique's ability to predict the presence of faulted components in the transmission of the SH-60B autonomously operated in a Helicopter Transmission Test Facility at Naval Air Warfare Center, Trenton, New Jersey. The analysis is implemented in the statistical software package S-plus (Mathsoft Inc., 1995).