Cellular proliferative activity has previously been determined by measuring the incorporation of radiolabelled nucleotides or by visual inspection of cellular morphology. Although two flow cytometric methods have recently been developed which can distinguish cycling from non-cycling cells, both have serious disadvantages. One method requires uptake of a substantial amount of BUdR, limiting its usefulness for in vitro systems. The other method utilizes RNA/DNA content differences but its successful application has proved cell-type dependent. We have now used the findings that the cell membrane is more highly polarized in resting than in proliferating cells and that cyanine dyes carrying a delocalized positive charge enter live cells to an extent that depends on the cell membrane potential, to develop a method of distinguishing between cycling and non-cycling cells. The greater the membrane polarization, the greater is the concentration of dye within the cell. At high concentrations, the dye molecules aggregate and their fluorescence is quenched. Thus, for a given external dye concentration, cells of different membrane potential would accumulate different amounts of fluorescent (non-aggregated) dye. Using fibroblasts in culture conditions chosen to provide various models of cycling and non-cycling cells, we found that fluorescence intensity with the dye, 3,3'-diheptyloxycarbocyanine (Di-O-C,(3)) was consistently greater in the former than the latter.