BACKGROUND: There is a growing recognition that common occupational injury surveillance systems in the US fail to reflect true injury risk; this failure limits efforts to accurately monitor efforts to prevent work-related injuries on a national level. METHODS: Data from the National Electronic Injury Surveillance System occupational supplement (NEISS-Work) were used to describe fall-related injuries treated in US emergency departments among workers in the construction industry (1998-2005). These data do not require workers' compensation as the payer in order to be classified as work-related. RESULTS: Based on NEISS-Work estimates, a total of 555,700 (95% confidence interval (CI): 390,700-720,800) non-fatal work-related injuries among workers in the construction industry were the result of a fall, resulting in an annual rate of 70 (95% CI: 49-91) per 10,000 full-time equivalents. Younger workers had higher rates of falls, whereas older workers were more likely to suffer serious injuries. The majority of the injuries (70%) were precipitated by falls to a lower level from roofs, ladders, and scaffolding. CONCLUSIONS: The patterns of fall-related injuries identified in these data are consistent with other reports. In contrast to the declining rates of falls requiring days away from work reported through the Bureau of Labor Statistics Survey of Occupational Injuries and Illnesses, construction industry fall-related injury rates estimated through NEISS-Work remained unchanged from 1998 to 2005 providing another perspective on this serious cause of morbidity in the construction industry.