Bovine respiratory disease (BRD) is one of the leading causes of morbidity and mortality in dairy calves. NAHMS (2011) reported that BRD affected 18.1% of pre-weaned heifer calves, with a mortality loss of 2.3%. The objective of this study is to establish a protocol for objective and efficient assessment of bovine respiratory disease (BRD) phenotypes in dairy calves and identify markers associated with BRD in a genome-wide association study (GWAS). 1,107 calves from 6 dairy farms in southern Wisconsin were measured at 3 and 6 weeks of age. Each calf was given a clinical score based on visual appraisal of eyes, nose, ears, attitude, cough, and temperature, as well as a subclinical score based on thoracic ultrasonography. The interaction of clinical and subclinical phenotype was represented on a 1 to 6 scale as overall BRD score. 1016 calves were genotyped with a commercially available single nucleotide polymorphism (SNP) array, and upon completion of quality control and imputation, 28,696 SNP and 1,014 individuals remained. A preliminary GWAS analysis was performed using a linear mixed model with SNP genotype as a fixed effect and with background polygenic effect as a random effect. Three- and 6-wk phenotypes were analyzed separately, and BRD scores were considered as binary (healthy or affected) or ordinal (6 levels reflecting increasing severity). At 3 wk of age, 8 and 6 significant SNP (P-value <5 × 10⁻⁵) were detected in the binary and ordinal analyses, respectively, with common SNP on chromosomes 1, 7, 17, and 18. At 6 wk of age, 3 significant SNP were found in each of the binary and ordinal analyses, with common SNP on chromosomes 8 and 9. Combining the clinical and subclinical scoring systems allows objective and efficient assessment of BRD for detection of important SNP using GWAS or whole-genomic selection against BRD. Further analysis is needed to identify putative genes affecting BRD and to assess the reliability of whole-genome predictions.