Membrane-potential-dependent accumulation of diS-C3(3) in intact yeast cells in suspension is accompanied by a red shift of the maximum of its fluorescence emission spectrum, lambda max, caused by a readily reversible probe binding to cell constituents. Membrane depolarization by external KCl (with or without valinomycin) or by ionophores causes a fast and reproducible blue shift. As the potential-reporting parameter, the lambda max shift is less affected by probe binding to cuvette walls and possible photobleaching than, for example, fluorescence intensity. The magnitude of the potential-dependent red lambda max shift depends on relative cell-to-probe concentration ratio, a maximum shift (572-->582 nm) being found in very thick suspensions and in cell lysates. The potential therefore has to be assessed at reasonably low cell (< or = 5 x 10(6) cells/ml) and probe (10(-7)M) concentrations at which a clearly defined relationship exists between the lambda max shift and the potential-dependent accumulation of the dye in the cells. The redistribution of the probe between the medium and yeast protoplasts takes about 5 min, but in intact cells it takes 10-30 min because the cell wall acts as a barrier, hampering probe penetration into the cells. The barrier properties of the cell wall correlate with its thickness: cells grown in 0.2% glucose (cell wall thickness 0.175 +/- 0.015 micron, n = 30) are stained much faster and the lambda max is more red-shifted than in cells grown in 2% glucose (cell wall thickness 0.260 +/- 0.043 micron, n = 44). At a suitable cell and probe concentration and under standard conditions, the lambda max shift of diS-C3(3) fluorescence provides reliable information on even fast changes in membrane potential in Saccharomyces cerevisiae.