Data collected from an 870-cow dairy herd using the Pregnancy-Specific Protein B (PSPB) serum ELISA for pregnancy diagnosis in lactating dairy cattle in place of rectal palpation. Data was compiled from 8 months of laboratory results and was compared to herd calving and breeding information, as well as rectal examinations. Due to the lag time between PSPB ELISA and rectal palpation, false negative classifications were unreliable; however the positive predictive value was valid and determined to be 95.2%. Two weeks of data were collected with minimal lag time and the sensitivity of the test from this data was found to be 95% and the specificity was 100%. The company's reported test accuracy was 99% sensitive and 91-95% specific. A cost-benefit analysis of the tests for this farm was performed taking into account the cost of the direct costs of the test as well as the cost of errors as would be expected to occur on the farm. The costs of each method of pregnancy diagnosis were very close and because of the large number of assumptions in the cost-benefit analysis it was determined that the cost difference between the two tests was likely negligible. Thus, the PSPB ELISA may be a convenient and cost-effective way for this farm to diagnose pregnancy in dairy cattle.