OBJECTIVE: To determine the specificity and sensitivity of a commercial copro‐antigen ELISA for the detection of Fasciola hepatica infection in cattle and sheep and to assess the suitability of the test for use in horses. METHODS: Testing was done on more than 100 negative faecal samples from each of sheep, cattle and horses and on at least 40 positive faecal samples from each species. Positive samples were selected based on a positive sedimentation test for liver fluke eggs. Faecal samples of animals from Western Australia, which is free of liver fluke infection, served as negative controls. Specificity and sensitivity were assessed for each species using the recommended kit cut‐off and also custom cut‐offs specific for each species based on the mean plus 3‐fold standard deviation of the mean of the negative samples for each species. RESULTS: Using the cut‐off recommended by the kit manufacturer, the specificity was 100% for all species and the sensitivity was 88%, 80% and 9% for sheep, cattle and horses, respectively. Using the lower custom cut‐offs for each species improved the sensitivity to 100% for sheep, 87% for cattle and 28% for horses, while maintaining the specificity above 99% for all species. CONCLUSIONS: The sensitivity of the commercial copro‐antigen ELISA can be improved by using custom cut‐off values for each species. With this modification, it is a suitable alternative screening test to the currently used sedimentation test for border control of sheep and cattle movement. The test is not suitable for use in horses.