The Faecal Egg Count Reduction Test (FECRT) is the most widely used method of assessing the efficacy of anthelmintics in horses. Equine Fae- cal Egg Count (FEC) data are frequently charatcerised by a low mean, high variability, small sample sizes and frequent zero observations. Accu- rate analysis of the data therefore depends on the use of an appropriate statistical technique. Analyses of simulated FECRT data by methods based on calculation of the empirical mean and variance, non-parametric boot- strapping, and Markov chain Monte Carlo (MCMC) are compared. The MCMC technique consistently outperformed the other techniques, indepen- dently of the sample size and distribution from which the data were gener- ated. Bootstrapping produced inappropriate 95% confidence intervals, con- taining the true parameter as little as 40% of the time, with sample sizes of less than 50, and tended to over-estimate the true FEC reduction rel- ative to the MCMC technique. Analysis of equine FECRT data yielded inconclusive results in 53 of 63 (84%) datasets, suggesting that the rou- tine use of prior sample size calculations should be adopted to ensure suf- ficient data is collected. The authors conclude that computationally inten- sive techniques such as MCMC should be used for analysis of FECRT data with sample sizes of less than 50, in order to avoid making erroneous in- ference of the true efficacy of anthelmintics in the field. Software to perform all three types of analyses documented here is freely available in the form of an add-on package to the R statistical programming language from http://cran.r-project.org/web/packages/bayescount/index.html.