The principal drawback of immunoassays for the determination of small drug molecules in biological fluids is their potential lack of specificity with respect to metabolites, However, the recent development of methods based on LC-MS/MS provides opportunities to test the specificity of radioimmunoassays (RIAs) by cross-validation. In this paper, we describe the use of a series of very simple statistics to test the equivalence of data obtained by both methods, and, thus, verification of the specificity of the immunoassay, We also describe the cross-validations of assays for three different drugs and show how each case provides a different insight into the specificity of the RIA that was being tested, We advocate the use of immunoassays to support human clinical programs provided, of course, that their specificity can be assured, We also explain the cost effectiveness of this strategy.