The aim of this thesis was to optimise the diagnosis of posterior circulation stroke (PCS), a less common stroke subtype with potentially devastating consequences to those patients with delayed or missed diagnosis. The first study investigated the frequency of delayed presentation and misdiagnosis of cerebellar infarction (CBI) in 115 consecutive patients presenting to the Emergency Department (ED) of a large comprehensive stroke centre. Most patients tended to present late (>4.5h, p=0.05). 34% were misdiagnosed, most commonly as having a benign peripheral vestibular disorder. Nausea and vomiting, the absence of neurological signs and isolated CBI significantly correlated with misdiagnosis (p<0.05). Only 10% of patients received acute intravenous thrombolysis and/or endovascular therapy. The study confirmed the need for an improved diagnostic paradigm to differentiate peripheral causes of dizziness and vertigo from PCS. The second study involving the same prospectively collected cohort demonstrated that patients with small CBI (<2cm, 33/108 [31%]) presented with fewer symptoms and signs compared to large CBI. Vertebrobasilar disease was the presumed aetiology in 37% of patients. Small CBI patients were less likely to have large vessel atheromatous vertebrobasilar disease. Cardioembolism affected 37% of patients, irrespective of CBI size or topography. The presence of moderate-severe small vessel white matter disease correlated with 3-month functional dependence, particularly in large CBI. The findings reinforced the importance of comprehensive vascular risk factor assessment including extensive cardioembolic work-up in patients with small CBI. The third study examined the diagnostic utility of non-invasive, quantitative vestibular function tests in differentiating between central (PCS) and peripheral (vestibular neuritis [VN]) causes of acute vestibular syndrome (AVS). Semicircular canal and otolith function tests were undertaken within 14 days of symptom onset in 22 patients with PCS and 22 with VN. Findings were compared to 40 normal controls. The study demonstrated that the vestibulo-ocular reflex gain and saccade characteristics obtained from video Head Impulse Testing (vHIT) were the most useful in separating VN and PCS. Vestibular function testing was also helpful in separating PCS from normal controls. These findings suggest that performing vHIT in AVS patients may help identification of the aetiology of AVS in emergency settings.