Background: Previous studies have shown that routinely completed free-text emergency department medical records contain limited information necessary for injury surveillance. We instituted an injury documentation sheet into our emergency department records to evaluate the impact on completeness of bicycle injury documentation rates. Methods: The pretest/posttest study design used E-codes to identify bicycle-related injuries. A standardized data collection tool was utilized to review these charts. Time periods before (January 1 to December 31, 2004) and after (January 1 to June 30, 2005) institution of a standardized documentation sheet were reviewed. Data were entered into the computer program, Epistat, and scores were used for comparison. Results: Initial review (n = 667) revealed mean age of patients 8.6 years, with 46% African American and 67% male. Helmet usage was documented in 49% of the charts (81 were wearing helmets; 245 were not wearing helmets). Mechanism of injury was documented as bicycle alone in 587, bicycle versus car in 13, and bicycle versus stationary object in 64.After implementation of an injury data sheet (n = 205), it was found that the mean age was 9.24 years, with 51% African American and 43% male. Helmet use was documented in 77% of cases (26 wearing helmets; 132 not wearing). Mechanism was documented as bicycle alone in 125, bicycle versus car in 66, and bicycle versus stationary object in 14. Helmet use was much more frequently documented after the initiation of an injury documentation reminder sheet (z = 6.97; P < 0.001; 95% confidence interval, 20.2-35.8). CONCLUSION:: The use of standard injury documentation prompts increased completeness of documentation. With improved documentation, more accurate injury surveillance can be performed. Copyright © 2008 by Lippincott Williams & Wilkins.