A descriptive study was undertaken on 595 dairy cattle abortion submissions to the California Veterinary Diagnostic Laboratory System from July 1, 1987, to December 31, 1989, to determine the etiologic nature and distribution (seasonal and geographical) of dairy cattle abortion in California as reflected by laboratory submissions. Univariate analysis was performed to characterize abortion-related submissions by farm and laboratory variables, and logistic regression analysis was performed to determine factors that may influence success of abortion diagnosis in the laboratory. The proportions of dairies that submitted abortion-related specimens from northern, central, and southern milksheds during the 2.5-year period were 20.3%, 15.7%, and 13.1%, respectively, and 60% of submissions were from medium-sized (200-999 cows) dairies. Submissions consisted of fetus (58%), placenta (2%), fetus and placenta (12%), and fetus, placenta, and maternal blood (0.84%); fetal tissues and uterine fluid constituted the rest. An apparent pattern in abortion submissions was indicated by a peak in submissions during the winter and summer of 1988 and 1989. Infectious agents were associated with 37.1% of submissions; noninfectious causes, 5.5%, and undetermined etiology, 57.3%. Bacterial abortion accounted for 18% of etiologic diagnoses; protozoal, 14.6%; viral, 3.2%; and fungal, 1.3%. Submissions comprising fetus, placenta, maternal blood, or their combinations were associated with a higher likelihood of definitive diagnosis for abortion than tissues, as were fresher specimens and submissions associated with the second trimester of fetal gestation.