Radioimmunotherapy (RIT) utilises antibodies directed against tumour associated antigens to carry a therapeutic dose of radiation to the tumour. Using RIT, model tumours have been successfully treated and yet clinical responses have been limited by poor tumour localisation. In an attempt to overcome this, many new antibodies have been developed. Measuring the gross tumour localisation and tumour to normal tissue ratio in animal models has generally been used to assess the potential clinical use of these antibodies. However, these measurements assume all the energy from the electron emitted from the radionuclide is deposited in the source organ, and also ignore the effects of dose-rate and cell proliferation during treatment. In addition, they do not consider the effects of heterogeneous dose deposition and response within the tissues. The principal purpose of this thesis is to develop a more accurate measure of the biological effect of radiolabelled antibodies in a mouse xenograft in order to select the optimal radionuclide/antibody combination for more effective therapy in man. A structural model has been developed from mouse data to facilitate more accurate absorbed dose calculations by accounting for organ size, shape, and position relative to surrounding organs. In addition, the linear-quadratic model, conventionally used in external beam radiotherapy, has been adapted for use in RIT to account for the effects of dose-rate and proliferation during treatment. To characterise heterogeneity of dose deposition and response in tumours, images of tumour morphology and radiolabelled antibody distribution were registered. The images were obtained through digitisation of stained histological sections and storage phosphor plate technology. All data was collected using a wide range of antibodies labelled with 131I and 90Y. These models show that multivalent, tumour-specific antibodies, with intermediate clearance rates, deliver the most effective dose to xenografts. Antibody affinity and avidity facilitate the prolonged retention in radiosensitive areas of tumour where most of the dose is deposited. In addition, a significantly greater activity of 131I can be injected before causing the equivalent bone marrow toxicity. Furthermore, when antibodies are labelled with 90Y, a significant amount of the electron energy escapes the source organ and is absorbed in surrounding tissue. Nevertheless, the results clearly show that radionuclide and antibody should be matched in order to deliver optimum therapy.