Johne’s disease (JD) is an infectious disease of ruminants caused by Mycobacterium avium subsp. paratuberculosis (MAP). JD is responsible for vast economic loses and some evidence links MAP to Crohn’s disease in humans. Control of JD is a priority for the dairy sector and relies on early identification of infected cows. Unfortunately, diagnostic assays are inaccurate at early stages of infection and to compensate for this low accuracy farmers usually test their cows repeatedly, classifying them as low or high-risk based on repeated test results. Farmers can then decide whether to breed from certain animals or not based on JD risk. Between 2008 and 2018, the two main UK milk recording companies tested almost 4 million milk samples from more than half a million cows. With support from the levy board that represents British famers (AHDB), we have shown that repeated test results can be combined to provide a more useful indication of the risk of a cow being MAP-infected. Based on that principle, we have developed a diagnostic algorithm that takes into account all test results from an animal as well as the dates of bovine tuberculosis testing and the probability that the cow was born infected. Pilot testing of this approach showed that if farmers make breeding decisions informed by our algorithm, the proportion of infected cows in the herd can be markedly reduced. In this project, we intend to progress the technology to a stage in which it is readily available for adoption by the industry.