This thesis describes in vitro experiments with the novel vascular targeting agent Combretastatin A4 Phosphate (CA4P) and non-invasive magnetic resonance imaging (MRI) measurements in patients treated with CA4P to derive parameters which reflect tumour and normal tissue blood flow and permeability. Shape changes induced following tubulin depolymerisation by CA4P are quantified in human umbilical vein endothelial cells (HUVECs) and are measurable after only 10 minutes exposure. The effect is more marked in proliferating than confluent HUVECs, and begins at doses that have no anti-proliferative activity. In contrast, human smooth muscle cells show no shape change after treatment. The similar time course of HUVEC shape changes in vitro and tumour vascular shutdown in vitro suggest that this might be an early event involved in vascular shutdown. The effects and recovery rates of several other tubulin-binding agents are compared with CA4P. Colchicine and vinblastine also induce changes in HUVEC shape but unlike CA4P, HUVECs do not recover after drug removal. For these drugs, shape change and antiproliferative effects occur at similar doses. ZD6126, which like CA4P also has vascular targeting activity at well tolerated doses, induces recoverable changes in HUVEC shape at doses with no anti-proliferative activity. The difference in recovery rates with different tubulin-binding agents might therefore be related to their therapeutic window. The reproducibility of dynamic contrast enhanced MRI is measured in 21 patients who had 2 pre treatment scans within a week. Comparing the technique in rats with an established method for measuring absolute blood flow provides verification that the kinetic parameters derived from this technique reflect blood flow changes. Significant reductions in these parameters in patients' tumours are seen 4 and 24 hours after treatment at well tolerated doses of CA4P at 52mg/m2 to 68 mg/m2. No significant mean changes are seen in kidney, liver, spleen or skeletal muscle, although a small proportion of patients have significant reductions which are generally not maintained, and not associated with clinical consequences.