A common fear is that air turbine handpieces suffer generalized corrosion of the bearings through being autoclaved. However, there are no studies to corroborate this for modern designs. Worthington et al. (1998) reported reductions in turbine speeds of 24-64% relating to clinical use and repeated autoclaving, while Leonard et al. (1999) observed that most (23 out of 25) handpiece bearing failures were due to failure of the (phenolic resin) bearing retainers (cages). Objectives: The objective of the present exploratory work was to observe and classify air turbine handpiece bearing failures occurring in actual clinical practice with a view to commenting on the possible relationships between autoclaving and bearing failure and to prepare for more detailed study. Methods: Forty-seven failed handpiece turbine assemblies (i.e. containing 94 bearing assemblies) from handpieces (KaVo 640C, Kavo 642B and NSK PanaAir) used in government general dentistry clinics were collected, disassembled and their bearing components observed using optical and scanning electron microscopy. Results: Defects observed were classified as cage failure, deposits on bearing components, wear, and corrosion. Overall, 27 bearing assemblies exhibited fragmented or melted bearing cages, 24 had marked soft surface deposits, 5 showed evidence of bearing ball wear and 64 had worn ball races. 48 showed some corrosion of the bearing races although heavy corrosion was present in only two of these. Corrosion of bearing balls was observed in 22 assemblies. Conclusions: Despite some scattered heavy corrosion, there was no evidence that generalized corrosion arising from autoclaving was the primary cause of failure. Most bearing failures appeared to have multifactorial origins. Irrespective of any possible role of corrosion in bearing failure, the need for proper infection control outweighs this consideration because effective sterilization of high-speed air-turbine handpieces before each patient use should be mandatory.