We describe an ELISPOT technique for the detection of antigen specific IFNgamma-producing T cells. The technique is performed on spleen cells plated directly ex vivo into ELISPOT trays without an in vitro pre-culture step. Thus, the assay is likely to reflect the in vivo activity of the cells. We have found that very high cell densities (at least 10(6) cells/well) are required for optimal detection of spot forming cells, and only at a high density of cells is the number of spots detected linearly related to the number of primed cells plated. If lower numbers of antigen primed cells are used, then unprimed spleen cells from syngeneic mice can be added to the well to increase the cell density. Under these conditions, we find that the number of spots is linearly proportional to the number of primed cells plated, even if these are well below a million cells/well. Experiments with MHC congenic mice indicate that the high density of spleen cells required to obtain optimal spot formation reflects a requirement for an MHC restricted function, probably efficient antigen presentation to T cells. The formation of IFNgamma spots is antigen dependent and abrogated by depleting the antigen primed cells of T cells. We conclude that this linear assay can be used to efficiently detect ex vivo antigen-specific IFNgamma-producing T cells.