The number of antigen-specific naive CD8(+) T cells is believed to be important in the shaping of adaptive immune responses, and is predictive for the magnitude of priming responses in mouse models. Because of extremely low precursor frequencies, knowledge about these cells comes from indirect techniques and estimations. Here, we present a strategy based on the combination of tetramer staining, magnetic-bead enrichment, and multiparametric cytometry, which permitted direct detection and analysis of CD8(+) T cells reactive for 6 different naive epitopes (MART-1(26-35), HIV-1 Gag p17(77-85), hepatitis C virus [HCV] NS3(1406-1415), HCV Core(132-140), NY-ESO-1(157-165), and cytomegalovirus [CMV] pp65(495-503)). Interestingly, we detected higher than 100-fold differences in precursor frequency across these epitopes (from 0.6 x 10(-6) to 1.3 x 10(-4)), but conserved frequencies among humans. Development of a procedure for direct assessment of T-cell precursor frequency in humans has important implications, with particular relevance to vaccine development and monitoring of tumor and self-reactive T cells.