Medical records of 33 dogs presented for acute onset, nonprogressive, intracranial dysfunction that had a magnetic resonance imaging diagnosis of brain infarction were reviewed. Postmortem confirmation of brain infarction was available in 10 dogs. All dogs were evaluated by CBC, serum biochemistry, thyroid and adrenal testing, urinalysis, thoracic and abdominal imaging, and cerebrospinal fluid analysis. Results of coagulation profile and arterial blood pressure were available in 32/33 and 28/33 dogs, respectively. On the basis of the imaging findings, infarcts were classified depending on their type (territorial or lacunar) and location within the brain (telencephalic, 10/33; thalamic/midbrain, 8/33; cerebellar, 15/33). No marked associations among location or type of infarct and patient age and sex, occurrence of systemic hypertension, and the presence or absence of a concurrent medical condition were identified. Small breed dogs (< or =15 kg) were significantly more likely to have territorial cerebellar infarcts, whereas large breed dogs (>15 kg) were significantly more likely to have lacunar thalamic or midbrain infarcts. A concurrent medical condition was detected in 18/33 dogs with brain infarcts, with chronic kidney disease (8/33) and hyperadrenocorticism (6/ 33) being most commonly encountered. Of 33 dogs, 10 were euthanized because of the severity and lack of improvement of their neurologic status or the severity of their concurrent medical condition. No association was identified between type or location of infarct and patient outcome. Dogs with concurrent medical conditions had significantly shorter survival times than those with no identifiable medical condition and were significantly more likely to suffer from recurrent neurologic signs because of subsequent infarcts.