Protein-based therapeutics have outstanding potential for the treatment of diseases, yet their physical and ' chemical protein instability has remained a major barrier for their rapid commercialisation. Development of efficient stabilisation protocols would be achieved by having a better understanding of the effect of key formulation parameters on protein stability. The main body of this project has been dedicated to investigating the effect of key formulation parameters, such as pH, ionic strength and temperature, on the stability and the aggregation mechanism of a model protein, Glucose oxidase at low (1 mg/ml) protein concentration. In addition, due to a strong trend in the pharmaceutical industry to formulate protein therapeutics at higher concentrations, key experiments were carried out at high (100 mg/ml) protein concentration. Changes in the physical state of the model protein were monitored using a combination of different techniques, including activity assay, Size-exclusion chromatography, visual assessment, intrinsic and extrinsic fluorescence spectroscopy, and static light scattering. In addition, we investigated whether the concept of accelerated stability study, widely used in the pharmaceutical industry, is a valid approach for predicting long-term stability. The results revealed that the rate and the mechanism of inactivation is pH, ionic strength, temperature, surfactant and protein concentration dependent. Furthermore, the results of the high concentration experiments showed the existence of an aggregation state, which appears to be different from the one observed at low protein concentrations. This is evident from the observation that · proteins at very high concentrations respond to formulations in a very different way than the same proteins at low concentrations. The findings from the work obtained from highly concentrated formulations showed the need for development of techniques that directly measure the stability of proteins at high concentrations without a dilution step. It also became evident that the concept of accelerated stability is not always applicable.