OBJECTIVES: A history of vascular disease has been shown to be associated with an increased risk of Parkinson's disease, but this may be due to reverse causation. In addition, non-steroidal anti-inflammatory drugs are thought to reduce Parkinson's disease risk, but evidence is inconsistent and confounding by prior vascular disease is possible. Our objective was to examine these associations in the UK prospective Million Women Study. METHODS: A total of 819,575 women were followed from median year 2001 for first hospital admission or death from Parkinson's disease. A history of ischaemic heart disease and of cerebrovascular disease was ascertained through hospital admission records and self-report at baseline. Use of the two non-steroidal anti-inflammatory drugs, aspirin and ibuprofen, was ascertained through self-report at baseline. Cox regression was used to estimate adjusted hazard ratios. RESULTS: During an average follow-up of 18 years, 10,364 cases of Parkinson's disease occurred. Women with prior ischaemic heart disease had an increased risk of Parkinson's disease (Hazard ratio 1.45, 95% confidence interval 1.35-1.54), as did women with prior cerebrovascular disease (1.51, 1.32-1.73); both associations were still pronounced 10 or more years after baseline. There was little association of aspirin use with risk of Parkinson's disease (1.06, 1.00-1.13) after adjustment for vascular disease, and no association with ibuprofen use (1.04, 0.97-1.11). INTERPRETATION: Ischaemic heart disease and cerebrovascular disease are both positively associated with subsequent risk of Parkinson's disease and these associations are unlikely to be solely due to reverse causation. In contrast, use of non-steroidal anti-inflammatory drugs is not materially associated with risk.