Photodynamic therapy of cancer exposes adjacent arteries to the risk of injury and the possibility of haemorrhage and thrombosis. The nature of photodynamic injury to normal arteries has not been satisfactorily defined, and the ability of arteries to recover with time is unclear. To clarify these issues, we have investigated the effects of PDT on rat femoral arteries, using a second-generation photosensitiser, disulphonated aluminium phthalocyanine, and a new method of photosensitisation, using endogenous synthesis of protoporphyrin IX following systemic administration of 5-aminolaevulinic acid (ALA). Pharmacokinetic studies of sensitiser fluorescence were carried out to determine peak levels of sensitiser. Subsequently photodynamic therapy at times corresponding to maximal fluorescence was performed using two light doses, 100 and 250 J cm-2. The nature of injury sustained and recovery over a 6 month period was investigated. Three days following PDT, all vessels treated showed complete loss of endothelium, with death of all medial smooth muscle cells, leaving an acellular flaccid artery wall. No vascular occlusion, haemorrhage or thrombosis was found. A striking feature was the lack of inflammatory response in the vessel wall at any time studied. Re-endothelialisation occurred in all vessels by 2 weeks. The phthalocyanine group showed repopulation of the media with smooth muscle cells to be almost complete by 3 months. However, the ALA group failed to redevelop a muscular wall and remained dilated at 6 months. Luminal cross-sectional area of the ALA-treated group was significantly greater than both control and phthalocyanine groups at 6 months. All vessels remained patent. This study indicates that arteries exposed to PDT are not at risk of catastrophic haemorrhage or occlusion, a finding that is of significance for both the local treatment of tumours and the use of PDT as an intraoperative adjunct to surgery for the ablation of microscopic residual malignant disease.