To investigate the hypothesized relationship between various daily atmospheric pollutant concentrations and hospital epistaxis presentation. A retrospective analysis of all presentations of non-traumatic epistaxis to our institution was performed, and the results were cross-referenced with London air quality data supplied by the National Air Quality Data Archive at the National Environmental Technology Centre (NETCEN) at the multi-centre tertiary referral ENT Department. An analysis of epistaxis presentation to St George's Hospital during the 5-year period January 1997-2002 was made from the patient administration system and accident and emergency records. The study includes 1373 emergency patients after traumatic and iatrogenic epistaxis were excluded. Factors measured were atmospheric ozone (O(3)), carbon monoxide (CO), sulphur dioxide (SO(2)), nitrogen dioxide (NO(2)) and particulate matter less than 10 mum in diameter (PM(10)). Hospital epistaxis presentation and atmospheric pollutant concentration were documented for each day of the 5-year period, weekly change in both parameters was recorded and statistical analysis was performed. The statistical methods used are linear correlation using Pearson's coefficient. Increased hospital epistaxis presentation was strongly associated with increased concentration of airborne particulate matter (r = 0.289, P < 0.001; significant if P < 0.05) and less strongly associated with increased concentrations of atmospheric O(3) (r = 0.150, P = 0.019; significant if P < 0.05). There was no association between epistaxis attendance and atmospheric concentrations of CO, NO(2) and SO(2) in this study. The concentrations of local airborne particulate matter and atmospheric O(3) in the days preceding hospital attendance for epistaxis are relevant when considering the associations of the episode. This further increases understanding of the pathophysiology of spontaneous epistaxis.