Purpose  To describe the difficulty of using Birmingham Eye Trauma Terminology (BETT) in epidemiological study, and introduce another mechanical eye injuries classification for epidemiological study based on BETT. 
Methods Setting: Multicenter investigation.
- Patient or Study Population: 10718 (11227 eyes) diagnosed as mechanical eye injuries were involved. From - January 2005 to December 2010, all medical records of eye injury inpatients admitted to 27 hospitals were reviewed.
- Intervention or Observation Procedure(s): All mechanical eye injuries were classified using BETT. Some eye injuries were difficult to classify using BETT.
- Main Outcome Measure(s): The injury type difficult to classify using BETT was recorded, so did the number of patients and percentage. 
Results Of 10718(11227 eyes) mechanical eye injuries, 1488 (1559 eyes) with merely orbital or ocular adnexa injury can not been classified, accounting for 13.88%. 1961 (2054 eyes) globe injuries associated orbital or ocular adnexa injury, accounting for 18.30%. 271 (284 eyes) with ocular surface foreign body (OSFB) or ocular wall foreign body (OWFB) can not been classified, accounting for 2.53%. 77 (89 eyes) with contusion and 9(11 eyes) with lamellar laceration associated OSFB or OWFB, accounting for 0.80%. 29(30 eyes) with rupture associated OSFB, OWFB or intraocular foreign body (IOFB), accounting for 0.27%. 60(62 eyes) with laceration associated OSFB or OWFB, accounting for 0.56%.
Conclusion It is difficult to classify some eye injury using BETT in epidemiological study and another mechanical eye injury classification for epidemiological study based on BETT is proposed.