Modern implantable medical devices use a variety of circuit and architectural features to ensure high levels of data integrity, especially data related to patient therapies and safety. Since modern implantable medical devices, such as implantable cardioverter defibrillators (ICDs) and pacemakers, rely on integrated-circuit (IC) technology, they are susceptible to soft errors from ionizing radiation sources. The IC technology employed in implantable devices typically uses a CMOS technology that is several generations behind the most advanced CMOS processes due to the overwhelming concern of transistor leakage, power consumption, and reliability requirements. For CMOS technology and modern circuit design processes, IC susceptibility to alpha-particle radiation is a known concern. For power consumption requirements, the operating voltages of implantable medical devices are also typically lower than standard electronic products, so this can have a much greater impact of alpha-induced soft-error rates (SERs). This paper will detail a case study performed on a new SRAM package, potentially for collecting diagnostic data, in which a variety of test methods was used to isolate materials contributing to alpha-induced soft errors. The test procedure utilized numerous standard SRAM components in a flip-chip configuration, with a test board to gather large amounts of data in a relatively short time period. SRAM test chips were also isolated from thermal neutrons by using a boron-shielded test chamber. Contributions from cosmic-ray-induced energetic neutrons were previously characterized and taken into account during the interpretation of the data. The correlation of soft errors to bump layout is shown with high-alpha-emitting materials, and the range of these errors is verified with modeling. The use of low-alpha-emitting solders and underfill materials for the reduction of soft errors will also be shown. Without underfill, residual errors from high-alpha-emitting solders, materials on the periphery of the circuit board, and the circuit board itself can be shown to impact the total SER of the system.