Fasciola hepatica is an important disease of livestock that is responsible for substantial economic losses worldwide. Estimates of the impact of infection on milk yield vary, likely reflecting different geographical locations, farm-level management, and diagnostic methods. Measuring anti-Fasciola antibodies on bulk tank milk (BTM) by ELISA provides a convenient herd-level diagnosis, but the utility of this test remains unclear. Therefore, we evaluated the utility of BTM ELISA test results in Danish organic dairy farms, including estimating the association between 305 day energy corrected milk yield (305d ECM) and F. hepatica infection both at individual and herd level. BTM samples from 218 organic farms were analysed using IDEXX ELISA and subsequently the farmers were interviewed during spring 2016 with the aim of characterising their management practices. The corresponding farm-level production data covering the period 2014–2017 were collected from the Danish national cattle registry. In the following year, 284 individual milk samples (4–7 per herd) along with BTM samples were collected from a subset of the same herds (n = 55). Linear mixed models were used to estimate the association between milk production and ELISA value at both individual and farm levels, and a generalised additive model was used to assess the relationship between within-herd prevalence and BTM ELISA. A dichotomised BTM result with positive outcome was associated with a reduction of 580.5 kg in average 305d ECM, and a positive outcome on individual-level ELISA was associated with a 919.5 kg reduction in milk yield for cows in their third or later lactations. A strong relationship between quantitative BTM ELISA sample to positive percentage (S/P%) and apparent within-herd prevalence based on dichotomised individual-level milk ELISA was also observed, although this relationship was non-linear in nature. We conclude that a useful indication of the within-herd prevalence of infection can be obtained from BTM ELISA following categorisation as negative, low, medium or high according to S/P% cut-offs of approximately 30, 80, and 150. This approach represents a cheap and useful diagnostic tool for monitoring the long-term success of control strategies for F. hepatica infections on a dairy farm.