Polymeric adhesives play an increasingly critical role in today's engineering designs. When used, adhesively bonded components reduce or eliminate the need for bolted or welded connections. In many cases, this can reduce stress concentrations and weight. With energy dissipating adhesives, noise and vibration reduction are possible, as is the use of unique or complicated designs that could not otherwise be constructed. Adhesive properties however, can vary greatly with time, temperature, and environmental exposure conditions such as moisture. It is therefore critical, to understand the behavior of adhesives over the range of conditions that a bonded component might experience. In this work, the behavior of a urethane-based adhesive was characterized and long-term durability predictions were developed as a result of the data collected. The popular T-peel sample geometry has been used extensively in this study to explore the mechanics of a bonded system and the resulting impact on adhesive durability. The T-peel specimens used, consist of two aluminum sheets or adherends bonded together, with tabs bent back in the shape of a "T" for gripping in a universal load frame. Unlike some other test geometries, T-peel samples are often made with relatively thin adherends that may experience significant plastic deformation during testing. This extraneous energy dissipation greatly complicates the analysis to extract meaningful fracture properties of the adhesive. During testing, the load required to propagate a crack in the adhesive layer is measured at fixed displacement rates. The total system energy can then be partitioned into the energy dissipated within the adhesive (fracture energy), and the energy dissipated through plastic work in bending of the adherends. By performing these tests at different temperatures and rates, the calculated fracture energies span a wide range of possible material behavior. Using the principles of Time Temperature Superposition (TTS), the collected data can be shifted to different times or temperatures. This behavior is well understood in polymer physics, and is made possible with material specific "shift factors". By using the principles of TTS, data collected in in a relatively short experimental window, can be used to accurately predict the behavior of the adhesive in years or even decades. In this work, nearly 200 T-peel samples were tested in four different studies. A preliminary set of unaged specimens was used to develop testing and data analysis methodologies. A second set of unaged samples was tested over a wide range of temperatures and rates, in addition to a third group, subjected to constant moisture and cyclically varying temperature. The final set of specimens, was exposed to 20 separate isothermal aging conditions. The experimental data showed that the 400+ cycles, were insufficient to statistically distinguish these samples from their unaged counterparts. Additionally, samples aged for up to 2000 hours in a dry environment, or 500 hours in a wet environment, showed no reduction in fracture energies in comparison with unaged samples. Specimens aged for more than 500 hours however, were observed to have a significant decrease in fracture energy values. Strong correlations between the thickness of the adhesive layer and estimated fracture energy values were found in this study. As adhesive thickness varied substantially due to manufacturing differences in the specimens tested, new analysis techniques were developed to deal with the variations in adhesive thickness. A MATLAB code based on the ICPeel program, was written to provide a spatial variation of parameters such as adhesive thickness, peel load, and fracture energy. This provided additional insights into the behavior of these T-peel coupons, and prompted the investigation of the Universal Peel Diagram concept. While this diagram was not found to be applicable to the adhesive tested in this study, the analysis indicated that T-peel coupons could be multivalued. That is, a single measured load value does not always describe an adhesive's fracture energy (as is widely believed). Depending on the sample's geometry and material properties, several measured loads could cause debonding. This has potentially far reaching implications on the selection of appropriate T-peel test geometries, as a single measured load is often assumed to correlate to an adhesive's true fracture energy. In this work, both aged and unaged T-peel specimens were tested and the basis of the Universal Peel Diagram investigated. Given sufficient exposure times to moisture, elevated temperatures were found to significantly reduce the amount of energy dissipated in the urethane-based adhesive. Additionally, the Universal Peel Diagram indicated that for some systems, the load required for debond is in fact, multivalued. Therefore, care should be taken when designing a T-peel test configuration to avoid the multivalued regions. / Master of Science / Polymeric adhesives play an increasingly critical role in today's engineering designs. When used, adhesively bonded components reduce or eliminate the need for bolted or welded connections, reducing their weight in the process. With adhesives, noise and vibration reduction are possible, as is the use of unique or complicated designs that could not otherwise be constructed. Adhesive properties, however, can vary greatly with time, temperature, and other environmental exposure conditions such as moisture. It is therefore critical to understand the behavior of adhesives over the range of conditions that a bonded component might experience. In this work, the behavior of a urethane-based adhesive was characterized in order to develop long-term durability predictions. Numerous test methods have been developed to characterize the behavior of adhesively bonded joints. In this work, T-peel specimens were used consisting of two aluminum sheets (the adherends), bonded together with tabs bent back in the shape of a "T" for gripping in a universal load frame. During testing, the load required to propagate a crack in the adhesive layer is measured. An outcome of this measurement and subsequent data analysis is the fracture energy—a measure of the effectiveness of the adhesive in transferring loads. If we perform these tests at different temperatures and loading rates, we can determine fracture energy values which span a wide range of possible material behavior. Using principles from basic polymer physics, the collected data can be shifted to different times or temperatures enabling us to accurately predict the behavior of the adhesive over years or even decades. In this work, nearly 200 T-peel samples were tested in four different studies. A preliminary set of unaged specimens was used to develop testing and data analysis methodologies. Unaged and cyclically (temperature) aged samples were tested over a wide range of temperatures and rates. The fourth set of specimens was subjected to 20 separate isothermal aging conditions and also tested at different temperatures and rates. The experimental data showed that the 400+ temperature cycles were insufficient to damage these samples significantly. Additionally, samples aged for up to 2000 hours in a dry environment, or 500 hours in a wet environment showed no reduction in performance in comparison with unaged samples. Specimens aged for more than 500 hours in a wet environment however, demonstrated a significant decreases in fracture energy values. Strong correlations between the thickness of the adhesive layer and estimated fracture energy values were found in this study, and new analysis techniques were developed to analyze the effect of these thickness variations on the joint performance.