Autoantibody tests are often ordered inappropriately. We aimed to evaluate the ordering patterns of these tests in our local health region and to develop a laboratory algorithm aimed at reducing unnecessary tests. Laboratory data including the number and sequence of tests, ordering physician specialties and results for antinuclear (ANA), extractable nuclear antigen (ENA) and anti-double stranded DNA (anti-dsDNA) antibody tests from 2007 to 2009 were evaluated. Based on this information and a clinical consensus meeting, an algorithm was developed and applied retrospectively to 1 year of inpatient laboratory data to simulate potential cost savings. We identified a large volume of these autoantibody tests performed, equating to testing costs of $862,706.72, where less than 17 % of each were positive. Repeated ANA tests were mostly ordered after a previously negative result, and 1 % of patients with negative results changed to ≥1:160 on repeat testing. Close to half of all ENA and anti-dsDNA tests that were ordered were done so simultaneously with ANA, suggesting their use as screening tests. This was done more frequently in the inpatient setting. An algorithm was developed where ENA and anti-dsDNA tests would be cancelled if ANA was negative in the same sample. ANA repeated within 1 year would be cancelled and the prior result provided. Application of the algorithm retrospectively simulated a 30 % cost savings. Repeat testing and simultaneous ordering of multiple tests contributed to the excessive ordering of autoantibody tests in our health region. Our proposed algorithm would reduce testing costs and should be accompanied by appropriate educational information for physicians.