Introduction:An arteriovenous malformation (AVM) is an abnormal collection of blood vessels in which arterial blood flows directly into the draining vein without the normal interposed capillaries. The Spetzler-Martin grading system has been widely accepted worldwide to estimate the pretreatment risks and predict the outcome of patients with intracranial AVM. In Malaysia, we still do not have the baseline data of this grading system. Methods: A total of 33 patients from a tertiary referral hospital diagnosed with intracranial AVM based on neuroimaging findings over a 4-year period were studied. Medical records were traced and neuroimaging findings were analysed. The AVMs were graded according to the Spetzler-Martin grading system and Fisher's exact test was used to assess statistical difference between the grades of the AVM and management plan for the patients. Results: Four patients were graded as Grade 1, 9 patients as Grade II, 10 patients as Grade III, 6 patients as Grade IV and 4 patients as grade V.Ten patients were treated conservatively; six patients underwent surgery and embolisationrespectively. Statistically significant difference (p=0.016) was found between the Spetzler-Martin grading system and the management of intracranial AVMs. Conclusion: The management decision was not made based on the grading of the AVMs. It is recommended that all AVM patients be routinely graded according to this system prior to treatment