Some of the more important statistical methods used in the analysis of experiments concerned with studies of phytotoxicity are described, and are illustrated by data from field and laboratory experiments undertaken within the Department of Agriculture of the University of Oxford. Most attention has been given to a consideration of quantal effects, such as the proportionate mortality. Adjustments to allow for natural mortality or the appearance of additional plants during the course of the experiment are outlined, together with the conditions under which the data should be transformed before analysis. Since the relationship between the proportionate response and some function of the dose or concentration of the toxicant generally follows a normal sigmoid law, the methods of probit analysis are appropriate for presize estimations. In this connexion, the design of experimets is discussed and the calculations involved in such an analysis are illustrated. In investigations where quantitative measurements are recorded, the dose-response relationehip may also be of the normal sigmoid form, so that the data be treated by a modification of the probit technique. The methods of statistical treatment demanded when the dose-response relationship does not conform to a normal sigmoid are briefly discussed.