The findings in this report are subject to at least two limitations. First, IA FACE surveillance, which involves more in-depth follow-up, only captured fatalities associated with work in Iowa. GPCAH surveillance, which is conducted in Iowa, Kansas, Missouri, and Nebraska, only captured accounts that appeared in newspapers or other periodicals. Therefore, reports from coroners or medical examiners, law enforcement, and emergency services were not obtained in Kansas, Missouri, or Nebraska. As a result, details about incidents in these three states often were limited (e.g., the age and sex of the decedent always were reported, but occasionally the decedent's activities and surroundings were not well reported). Second, reliance primarily on news reports means that some fatalities might go unreported. In Iowa, during 2003-2007, all seven of the fatalities caused by cattle that were documented by the state-based Census of Fatal Occupational Injuries (CFOI) of the U.S. Department of Labor's Bureau of Labor statistics also were captured through IA FACE and GPCAH surveillance. However, CFOI documented four cattle-caused fatalities in Kansas, seven in Missouri, and four in Nebraska, whereas GPCAH captured only two fatalities in Kansas, six in Missouri, and two in Nebraska. These data indicate that in states where only press clips were used to document agricultural fatalities, five out of 15 (33%) of the fatalities were unreported, suggesting a sensitivity of 67%. However, the advantage of using press reports is that more information regarding the circumstances of the deaths might be collected. In published studies, the sensitivity of newspapers as an injury surveillance source has varied according to the type of injury.9