Specifying the molecular basis and clinical significance of cluster formation between antigen-presenting cells and T lymphocytes will be important in many areas of immunology. In this paper we describe a novel and reproducible technique for measuring cluster formation in suspension between purified human blood monocytes and purified autologous T lymphocytes, and its application to determining the effects of recall antigens and mitogen. Blood monocytes and T lymphocytes from eight normal subjects were separately prelabelled with two different carbocyanine dyes prior to co-culture in suspension with or without antigen (PPD, SKSD) or mitogen (PHA). At 24 h the co-cultures were examined for cluster formation by ultraviolet microscopy and flow cytometry. Control experiments showed that the carbocyanine dyes were non-toxic in vitro, that cell labelling was stable for culture periods up to 120 h, and that the two dyes did not leak from cell to cell. By this technique we measured the proportion of monocytes clustering one or more T lymphocytes in the presence and absence of recall antigen or PHA. There was a close correlation between visual and flow cytometric measurement of monocyte: T lymphocyte clustering ( p < 0.001) as well as a close relationship between the ability of the two recall antigens to increase the extent of clustering above the baseline ( p < 0.001). Antigen-increased cluster formation did not correlate with baseline clustering, unlike PHA-increased clustering, which was related to baseline levels ( p = 0.02), suggesting the operation of distinct mechanisms. The method is applicable to measuring cell-cell associations in suspension during extended periods of culture, as well as for the study of agents which might modify intercellular adhesion processes.