Purpose: 1/3 of all diagnoses of epilepsy are made in patients over 60 (Tallis et al, Age and Aging, 1991; 20: 442-448) and this diagnosis is often termed late onset epilepsy (LOE). When no clear underlying aetiology is found occult cerebrovascular disease (CVD) is often implicated. The aim of this study was to investigate the association between LOE and structural and functional cerebrovascular markers using magnetic resonance imaging (MRI). Method: MRI scans were performed using a 3T Philips whole-body scanner. The imaging protocol included a T1 weighted image - GM volume; FLAIR sequence - WML lesions and MRI-ASL for CBF and tA. A non-rebreathing circuit was used in conjunction with a gas mixture (21% O2 and 79% CO2), to induce hypercapnia in subjects during scanning. Baseline CBF and tA values were extracted from the whole brain region during the initial 5 min breathing room air. The vasodilator stimulus of hypercapnia enabled calculation of cerebrovascular Reactivity (CVR) as a% change in CBF or tA divided by the% change in ETCO2. Results: 15 HC and 14 patients with LOE participated in the study. Patients with LOE had significantly lower GM volume than HC (mean [SD]: 0.34 [0.04] vs. 0.38 [0.01], p = 0.02) and a trend towards a higher WMLvolume than HC (mean [SD] 1416.5 [1498.6] vs. 514.2 [480.9]. Baseline CBF did not significantly differ between the two groups but baseline tA was found to be significantly longer in patients with LOE than HC (mean [SD] 1538.9 [128.5] vs. 1363.1 [166.6], p <0.01). Measures of CVR both in CBF (p = 0.85) and tA (p = 0.51) were found not to differ significantly between the two groups. Conclusion: Differences in structural and functional cerebrovascular markers lend support to the concept that occult CVD may be important in epileptogenesis.