Objective To assess the prevalence of tooth wear in adolescents and its relationship with diet, saliva and gastro-oesophageal reflux. Design Single centre cluster sample epidemiological study. Setting A school in London in the summer of 1996. Subjects 11-14-year-old schoolchildren. Main outcome measures The Smith and Knight tooth wear index (TWI), salivary factors, diet and symptoms of gastrooesophageal reflux were recorded for all subjects. Results Results were obtained from 210 subjects. One subject refused to provide a saliva sample and 11 subjects provided insufficient saliva for analysis of buffering power (n = 198). 57% (95% confidence intervals 50.3-63.7%) of subjects had tooth wear on more than ten teeth and a median 12% (interquartile range 6-18%, 95% confidence intervals 8-14%) of surfaces were affected. However, dentine involvement was rare. The median intake of carbonated drinks was 2 cans (interquartile range 1-3) a day. However, there was no correlation with TWI (r = -0.09, P = 0.19). There was no relationship between tooth wear index (TWI) and salivary flow rate (r = -0.02, P = 0.78) or buffering capacity (r = -0.02, P = 0.76). A trend was observed for those with a reported history of regurgitation (n = 27) to have a higher maxillary TWI (median 8, interquartile range 2-13) compared with those who did not (5, 2-9, P = 0.06). Conclusions Tooth wear is common in adolescents and the relationship with dietary acid, salivary buffering and symptoms of gastro-oesophageal reflux is complex and requires further investigation.