The objective of this study was to characterize genetic, environmental, and economic factors related to the incidence of bovine respiratory disease (BRD) in feedlot calves. Records from 18,112 calves representing 9 breeds (Angus, Braunvieh, Charolais, Gelbvieh, Hereford, Limousin, Pinzgauer, Red Poll, and Simmental) and 3 composite types (MARC I, MARC II, and MARC III) over a 15-yr period (1987 to 2001) were evaluated. Disease incidence was observed and recorded by station veterinary and technical staff. The incidence of BRD varied across years, with the annual observed incidence ranging from 5 to 44%. From 1987 to 1992, the annual average incidence generally exceeded 20%. However, in later years the annual incidence did not exceed 14%. The epidemiological pattern indicated that BRD infection increased dramatically after 5 d on feed and remained high until approximately 80 d on feed. Previous BRD infection during the preweaning period did not influence subsequent BRD infection in the feedlot. Steers were more likely to become sick with BRD than heifers; castration before entry in the feedlot may be a predisposing cause. Few significant differences among breeds were detected for BRD incidence. Adjusted solutions from mixed model analyses indicated that Herefords were generally more susceptible to BRD infection (P < 0.05) than MARC I and III composite types. Composite breed types had similar susceptibility compared with other purebred breeds. Mortality associated with BRD was greatest in Red Poll calves (9%) compared with the average over all breeds (4%). Estimates of heritability for resistance to BRD ranged from 0.04 to 0.08 +/- 0.01. When the observed heritability was transformed to an underlying continuous scale, the estimate increased to 0.18. Selection for resistance to BRD could be effective if phenotypes for BRD resistance were known. Thus, development of an inexpensive and humane method of challenging animals with BRD to determine resistance would be an important step in reducing the incidence of BRD. This study also demonstrated that producer-collected field data could be used for selection against this disease. The economic loss associated with lower gains and treatment costs for BRD infection in a 1,000-cattle feedlot was estimated as dollar 13.90 per animal, not including labor and associated handling costs.