Objective: Describe the characteristics of patients admitted to the otolaryngology emergency department for an episode of epistaxis and analyze epidemiological, clinical, endoscopic data between the two groups in order to establish a model that predicts the risk factors associated to epistaxis. Methods: prospective study including 161 adult patients spread over a month in the department of otolaryngology; interesting two groups; patients who had epistaxis (group 1) and those who had another emergency (control group = group 2) in the same study period .A comparison was made of age, sex, time of consultation, surgical or traumatic history, taking medication (anticoagulants, nasal corticoids), hemodynamic parameters at admission (blood pressure, pulse, hemoglobin). Intergroup variations were analyzed using t student and chi-square tests. Univariate and multivariate logistic regression were used to establish a predictive model of occurrence of epistaxis. Results: Univariate logistic regression confirmed that the occurrence of epistaxis was associated with male gender, consultation during the night, an history of High Blood pressure (BP) or Coagulopathy, a high Mean systolic BP, a high Mean diastolic BP, high Cardiac frequency or a low Hemoglobin level at admission. According to the multivariate logistic analysis, the predictive factors of epistaxis despite were: male gender (OR =0.089, 95% CI [0.008-0.978], p = 0.048), history of high blood pressure (OR = 0.009, 95% CI [0.00-0.236], p = 0.005). Furthermore, the multivariate analysis excluded the effect of age, history of cardiovascular disease, or anticoagulants. Conclusion: The only predictive factors of epistaxis were male gender, and arterial Hypertension.