The primed lymphocyte test (PLT) has been used as a method to identify HLA-D or HLA-D-related (HLA-DRw) specificities; however, the assignment of HLA-D or HLA-DRw specificities by this procedure is often nondiscriminatory with false positive and false negative results. We have investigated three approaches to improve PLT discrimination: 1) restimulation of primed cells after various times in primary culture; 2) restimulation using a small number of responding PLT cells per well, and 3) nylon wool purification of responder cells before initial priming. The results show that when PLT cells are restimulated and the response is plotted on a log/log scale in relation to a decreasing number of responding cells per well, a characteristic “linear” relationship is seen when less than 1.0 × 104 PLT cells are used (stimulators = 5.0 × 104 cells per well). When greater than 1.0 × 104 PLT cells per well are used, a linear response is seen only when the priming culture is prolonged to 17 days or more. It is only when the PLT cell is in this “linear phase” of response that the responses to various stimulating cells can be compared to yield the most discriminatory % relative response (RR). Finally, nylon wool treatment of responder cells before sensitization improves subsequent discrimination in PLT when large numbers of primed responders per well are used. Thus, optimum discrimination in PLT is obtained: 1) by using a low number (less than 1.0 × 104) responding PLT cells per well, 2) by prolonging the priming MLC to at least 17 days, and 3) by pretreating the responder cells with nylon wool. These experiments suggest that during allogeneic stimulation in MLC, there is activation of a population of lymphocytes that react with HLA-D or HLA-DRw determinants and that other nonactivated or unprimed cells within the PLT population may alter a discriminatory response; such cells must be functionally decreased by these procedures for optimum results.