In his article in last month's issue of TiPS (14, 89-91), Douglas Craig discussed the application of the Cheng-Prusoff relationship to the analysis of antagonists in functional experiments. Craig described how the translation of this method from biochemistry to pharmacology can be accompanied by errors due to theoretical misunderstanding. He also emphasized the lower level of rigour associated with this means of estimating antagonist affinity constants compared with Schild analysis. In this article, Paul Leff and Iain Dougall extend Craig's arguments by describing some additional ways in which the application of the Cheng-Prusoff method to pharmacological experiments are likely to result in misinterpretation and erroneous estimates of antagonist affinity.