I read the article by Brzozowski et al. about the percentage of intracranial haemorrhage in type I dural arteriovenous fistula with great interest [1]. In the paper, the authors demonstrated that 50% of patients with dural arteriovenous fistula presented with intracranial haemorrhage. However, the results of the paper were not concordant with the already published data in some larger studies (n = 236-1075), which have demonstrated a risk of intracranial haemorrhage between 12% and 23.7% [2-5] (Table 1). In the paper published by Li et al., only two cases out of the reported 91 with Borden type I (Cognard type I and type IIa) dural arteriovenous fistulas presented with an intracranial haemorrhage, as compared to 28.6% in the authors’ paper. The sample size of the study and the inclusion of only those patients who were managed with embolisation probably contributed to this discrepancy.