Fasciola hepatica (the liver fluke) is a common, global parasite of livestock. It can be highly pathogenic and has health and welfare implications for infected individuals. Typically, in ruminants, infections are sub-clinical, but if undiagnosed, they can lead to significant production losses. Accurate diagnosis is crucial to identify infection. Antibody detection ELISAs are commonly used to diagnose infection due to their high sensitivity and specificity and are typically based on native fluke excretory/secretory (ES) products or cathepsin L1 (CL1), the immunodominant antigen within ES products. These tests have been developed based on the antibody response of experimentally infected animals; however, this response has not been well characterised in naturally infected animals. We compared the antibody recognition of a recombinant CL1 (rCL1) antigen and native adult fluke ES products. Whilst samples from experimentally infected animals showed strong recognition of rCL1, serum antibodies from naturally infected animals did not. These results were confirmed by peptide array. Immunoblotting sera against ES products showed that experimentally infected animals had a strong, specific response to CL1/CL2 proteins whilst antibodies from naturally infected animals recognised multiple proteins and had a variable response to CL1/CL2. Mass spectrometry of proteins separated by 2D SDS PAGE, identified several antigens recognised by serum antibodies from a naturally infected cow, including cathepsins L1, L2 and L5, glutathione S-transferase and a dihydrolipoyl dehydrogenase. Overall, these results show that the antibody response in naturally infected animals to adult fluke ES products is qualitatively different to experimentally infected animals. This suggests that a diagnostic test based on CL1 alone may not be appropriate for diagnosis of natural F. hepatica infections in sheep and cattle.