However by 72 h post-stroke, vascular density appeared to have recovered in Nox2 KO mice, indicating that while Nox2 deletion does not confer long term protection on neurons in the core, it may be beneficial in terms of restoring the cerebral vasculature to pre-stroke levels. It remains unclear whether the absence of Nox2 activity within the vessels themselves is beneficial, or if the absence of Nox2 from surrounding cells creates an environment more conducive to new vessel growth. Compensatory changes in other NADPH oxidases may also be involved in the observed effects, with Nox4 reported to play a significant role in angiogenesis. Angiogenesis is thought to be essential for ischaemic brain repair as it stimulates blood flow and metabolism in the ischaemic boundary, and may provide critical neurovascular substrates for neuronal remodelling. Proliferating endothelial cells, indicative of angiogenesis, have been shown to increase as early as 24 h after stroke in a mouse MCAo model. We have previously shown the importance of Nox2 in LEE011 promoting cell survival during rat endothelial cell proliferation in vitro and that inhibition of Nox2 significantly suppresses vessel growth. Others have reported that perfusion recovery and an increase in capillary density are significantly inhibited in Nox2 KO mice at 7–14 days after ischaemic hind limb injury. The current study is the first to show that Nox2 deletion does not affect vascular loss after stroke but does result in increased vascular staining in the damaged brain by 3 days. This apparent beneficial effect of Nox2 deletion on vascular recovery may reflect the early time point examined. Alternatively this may be a lasting effect, reflecting the differential expression and roles of the NADPH oxidases in the cerebral vs. systemic vasculature. There is a substantial body of research which suggests that blocking the production of ROS by targeting a responsible enzyme may deliver a more favourable therapeutic outcome after ischaemic stroke with reperfusion. The NADPH oxidases present such a target, however the transient nature of the protection afforded by Nox2 deletion in the current study suggests that other factors may be of greater significance in the search for acute therapies. While there is evidence to show that Nox2 is harmful in the acute phase of stroke, the results of the present and previous studies suggest that it may merely delay infarct progression, and not prevent it. Of particular interest in the current study is the potential to target Nox2 for promoting early re-vascularisation after stroke. Therapeutics aimed at manipulating Nox2 may improve brain repair and subsequent long term functional recovery, the ultimate goal of any stroke therapy. Persistent oxidative stress is suspected to be an underlying factor in many of these degenerative effects observed at the organ level.