Several methods exist to estimate animal-level prevalence of Johne's disease from pooled faecal samples. A recently developed user-friendly computer program - the Pooled Prevalence Calculator - enables ready calculation of animal-level prevalence by seven methods, including methods for imperfect test sensitivity and specificity and for variable pool size. This paper will compare prevalence estimates and confidence intervals between methods and provide comment on the limitations of each method for use in field based research. Prevalence estimates presented in this paper result from analysis of pooled faecal culture (PFC) data from the field trial of a killed vaccine for the control of ovine Johne's disease (OJD), undertaken on three farms in New South Wales, Australia. Estimates for one observation per farm at approximately 30 months post-vaccination were compared between methods and with the results of individual faecal culture. These results illustrate: the trend in prevalence estimates calculated by methods that assume PFC is a perfect test versus an imperfect test the inability to apply frequentist methods that assume an imperfect test when OJD prevalence is high the impact on the Bayesian method of an inappropriate prior estimate of prevalence. To conclude we will provide guidelines on the more appropriate methods for prevalence estimation on data such as that from the vaccination trial and provide comment on further work required to support prevalence estimation from pooled samples in Johne's disease research.