Agriculture has consistently been identified as one of the most hazardous occupations in the United States, with rates of morbidity and mortality more than twice those for all occupations. Currently, no systems exist that can provide accurate data on the incidence of, and risk factors for, agricultural injury. To date, the majority of studies concerning agricultural injury have used case-series or hospital and emergency department data rather than population-based data. Although case series data are important for descriptive purposes and identifying problem areas, such cases account for only a small proportion of the injury problem in agriculture. The need for injury data systems and the use of epidemiologic data as the basis for developing appropriate injury intervention strategies have been reported as essential to the ultimate control of this major public health problem. The objectives for this study were to: 1) identify risk factors for farming/ranching operation related injuries to persons 20 years of age, using a case-control study design; 2) determine the incidence, types, sources, severity, and social and economic consequences of injuries by using an injury data collection system that can serve as a basis for surveillance; and 3) modify the RRISI/pilot study rural population injury surveillance data collection system instruments for the current effort, thus, enabling its transportability to other geographic locations nationwide. The relevant research design and specially designed data collection instruments enabled accomplishment of these objectives. The study involved a cohort of farming/ranching operation households in Minnesota, Wisconsin, North Dakota, South Dakota, and Nebraska. Data were collected for the two six-month periods of 1999 to identify all injury events and relevant demographics for all household members; data pertinent to numerous exposures of interest were collected for children and youth, <20 years of age, through the application of a simultaneous nested case-control study. A random sample of 3,200 operations was selected for each state (total n=16,000), from the United States Department of Agriculture's (USDA) National Agricultural Statistics Service (NASS) Master List Frame. Introductory letters were sent to each operation; subsequent screening telephone interviews were administered, using a computer assisted telephone interview (CATI). Eligibility involved: being actively engaged in farming/ranching as of January 1, 1999; having sales of agricultural goods >$1,000 in the past year and/or land registered in the Conservation Reserve Program [CRP]); and having a household associated with the operation that included at least one child <20 years of age, as of January 1, 1999. Each eligible household, that agreed to participate, subsequently received packets containing detailed information and specially designed cards to assist them in the two subsequent full data-collection interviews. Any injurious event, that met one or more of the following criteria, was included: restricted normal activities for at least four hours; resulted in loss of consciousness, loss of awareness, or amnesia for any length of time; or, required professional health care. Agricultural-related injuries were those that resulted from any activity related to an agricultural operation, or occurred as a result of bystanding in relevant areas. To determine the total injury burden on the agricultural population, data on injury events related to agricultural operation activity and all other activities, were collected. For the case-control study, cases were those who incurred an agricultural-related injury associated with their operation; up to six controls, per case, were sampled from the population at risk. Interviews enabled data collection on exposures of interest during the months prior to the injury events for cases, or during the months randomly selected for controls, based on an injury incidence algorithm. Validation, relevant to selection bias and information bias, was incorporated. Personal risk and injury event rates were adjusted for within-household correlation using generalized estimating equations (GEEs), excluding levels for missing values and non-response. Potential selection bias was controlled by inversely weighting observed responses with probabilities of non-response, estimated as a function of characteristics available from the NASS database. To account for unknown eligibility among non-respondents, probability of eligibility was estimated from these same characteristics and also used to weight responses. Analyses of the case-control study included both univariate and multivariate; based on the causal model and relevant directed acyclic graphs, variables were selected to enter in the multivariate model analyses. Logistic regression was used to investigate the relation between specific exposures of interest and the occurrence of agricultural-related injuries. A total of 16,538 persons were followed through the study period; 51 % were <20 years of age. A total of 2,586 total injury events were reported for the study population; 1,198 (46.9%) occurred on one's own agricultural operation; 68 (2.7%) on someone else's operation; and 1,291 (50.5%) were related to activities other than agriculture. Respective rates for these classifications were 74.6,4.3, and 81.4 injury events per 1,000 persons. The overall annualized rate of injury was only 1.2 times greater for those 20+, compared with <20 years of age (176.0; 145.9). Based on multivariate analyses, the odds of sustaining an injury increased as the number of hours worked per week on one's own operation increased. The primary sources of injuries, associated with farming/ranching for those <20 years, were animals (41 %) and falls (31 %); for those 20+ years, they were also important sources, as were machinery (19%) and tractors (13%). Consequences of the agricultural-related injury events, for those <20 and 20+ years, respectively, included: treatment by a health care professional (79%; 82%); restricted activity for >4 hours (77%; 71 %); and hospitalization (4%; 5%). Restriction from regular activities for >7 days was reported for 29% of each age group. Of further interest is the impact of injuries, both agricultural-related and those associated with other activities, upon the farming operation; 17% and 14%, respectively, of those <20 and 20+ years of age, identified >7 days of lost agricultural work time, while for non-agricultural-related injuries, this accounted for 17% of each age group. Based on multivariate analyses of case-control data, involving those <20 years of age, increased risks were identified for operating or riding in a motor vehicle and riding on or operating a tractor; increased risks for operating either large or small equipment were suggestive. For animal exposures, increased risks were identified for working with horses, sheep, and beef cattle; exposures to swine and dairy cattle were also suggestive of risk. This effort has enabled identification of the incidence and consequences of agricultural injuries, in concert with the burden of all injuries, on the agricultural operation for all persons, and the risk factors for agricultural-related injuries among persons less than 20 years of age. Most importantly, the latter data serve as a basis for development of prevention and control strategies essential for the reduction of morbidity and mortality from injuries incurred by children as a result of agricultural operation activities.