A variety of graphical and computational methods can be used to derive a median lethal concentration (LC50) from concentration-mortality data produced by an acute mortality test. In the selection of a method, practical considerations should receive as much attention as the usual theoretical and statistical considerations. No method should be used that does not calculate both the LC50 and its 95 percent confidence limits. Unfortunately, several computational methods will not calculate an LC50 from some concentration-mortality data that are practically and statistically useful to aquatic toxicologists. Useful, statistically sound information about the LC50 and its 95 percent confidence limits can be calculated from the data produced by any acute mortality test that meets published criteria of acceptability, by using an internally consistent scheme based on the moving average method and the binomial test.