The body routinely encounters new micro-organisms and epithelial surfaces provide the first line of defence. Most micro-organisms that get passed this barrier are detected and eliminated by the immune system. The liver plays an important role. Patients suffering from acute liver failure (ALF) are more prone to bacterial infections, and death from sepsis is not uncommon. Liver transplantation, is a successful therapy available for ALF, but too few donor organs are available. A bio-artificial liver (BAL) machine could “buy time” for a patient to recover or to bridge to eventual transplantation. An extracorporeal device was developed that can be integrated into the BAL, to treat in real time plasma infections, eliminating pathogens whilst conserving the integrity and functionality of plasma components, as one component of the BAL. Our hypothesis was that phototreatment is the most suitable pathogen reduction technology (PRT) for real time applications. The efficacy of a number of photosensitisers and light sources were evaluated. Custom equipment was developed to investigate the characteristics of semiconductor light sources. Their optimum operating conditions were investigated in terms of operating temperature, supply current and placement. Further experiments were performed to select appropriate materials for the construction of the phototreatment exposure unit. A first PRT prototype was developed and used to illuminate samples of E. coli in saline solution. The experiments demonstrated more than 4 log 10 reduction of surviving colony formation units, thus proving the method antimicrobial. A second PRT prototype was developed and used with a wider choice of light intensity, wavelengths and photosensitisers. It was used to perform further phototreatment experiments against E. coli, S. warneri and C. albicans. Significant reduction of pathogens was demonstrated in all cases. This thesis has demonstrated the initial proof of this hypothesis, leading to further in- vestigations being performed and the development of a clinical scale model.