In calculating the effectiveness of the latest UK badger-culling targets for controlling bovine tuberculosis, Christl Donnelly and Rosie Woodroffe do not consider the uncertainties in estimating badger populations or how information collected during culling is used to evaluate the success of culls in real time (Nature 526, 640; 2015).

Experience shows that there is greater uncertainty associated with badger population estimates than previously thought. A post-cull assessment by the Department for Environment, Food and Rural Affairs (DEFRA), using the number of badgers removed and reductions in sett occupancy, suggests that badger abundance may have been overestimated. Using the mean of the population estimate to establish a minimum number to be culled, as implied in Donnelly and Woodroffe's calculations, leads to a high probability of a culling objective that could greatly exceed actual badger numbers.

The current culls use methods similar to a trial that ran from 1998 to 2006 in southwest England and the west Midlands. The trial achieved a roughly 70% reduction in badgers, with large variance between trial zones, based on post-hoc assessments. Applying similar culling effort to a zone should converge on a similar outcome to the trial.

To reduce the badger population by a similar proportion as in the trial while providing an achievable objective, the government has set an initial minimum culling number at the lower end of the estimated population range. Information gathered during the cull will be used to assess whether this number should be increased. The most up-to-date data about the badger population are used to assess whether the culls are removing enough badgers and are therefore likely to achieve a similar outcome to the trial.

Comparison with control zones, where there has been no culling, provides no indication that culling has increased disease in cattle, as was widely predicted in advance of the culls (see go.nature.com/grk4ri).