Numerous biochemical datas support the noxious role of anti-inflammatory drugs on immune response. Those observations are often put forward for unfavorable evolution of odontogenic infection but has never been really proven in clinic. The aim of this study is to try to clarify this role based on the collection of the clinical course of odontogenic infections over a 10-year analysis period. The investigators implemented a prospective observational study. The sample was composed of patients managed between January 2004 and December 2014 for severe odontogenic infection based on three criteria: hospital admission, intravenous antibiotic therapy, tooth extraction and collections drainage under general anesthesia. Clinical and pharmacological data were collected at admission, during hospitalization until discharged home. The population was first separated into two groups patients with or without anti-inflammatory drugs on admission, then on four groups (non-steroidal anti-inflammatory drugs, corticosteroids drugs, both and none on admission). Analysis were performed each time by univariate analysis, multivariate analysis and propensity score matching. Six hundred and fifty-three patients were included in the study, 329 (50%) patients report orally anti-inflammatory treatment before presenting to hospital, 50 (7.6%) received corticosteroids, 242 (37%) received NSAIDs and 37 (5.6%) both. Evolution is worsening for patients under anti-inflammatory drugs in term of hospitalization in ICU (p=0.016), number of surgeries (p=0.003), risk of tracheotomy (p=0.036), duration of hospitalization (p=0.005) and spaces involved by the infection (p<0.001). When separating patients into 4 groups, dysphonia and odynophagia are more frequent for patients under corticosteroid and NSAID (35.14%, p<0.001), mediastinal erythema is more frequent for patients under corticosteroid (16%, p=0.004), fever is more frequent for patients under NSAID (35.5%, p=0.032), pain is higher for patients under corticosteroids (p=0.024). But, in order to reduce bias, linked to factors of gravity, a regression weighted by propensity scores was performed and any group of patients is different from the others. Patients under anti-inflammatory drugs have more severe dental infection on admission and their complex evolution seems to be linked to the severity of infection on admission.