AbstractThe Farm‐Scale Evaluations (FSE) is a three‐year experiment to study the effects on farmland wildlife of the commercial management of four genetically modified herbicide‐tolerant (GMHT) crops—beet, maize, spring oil‐seed rape and winter oil‐seed rape. The experiment records abundance of a wide range of weed and invertebrates in the arable system. The FSE is a paired comparison arranged as a randomized block design with two treatments (GMHT and conventional crop management) and fields as blocks. The choice of the half‐field as the experimental unit is based on field availability, the primary focus on the vegetation, the need to reduce variability and gains in sampling effort. Farms and fields are selected to reflect the range of geographical variation and farming intensity for each crop. A power analysis shows that using about 60 fields per crop over the three years will be sufficient to detect differences of ecological significance. Copyright © 2003 John Wiley & Sons, Ltd.