The methylxanthine, caffeine, quenches the fluorescence of the ratiometric Ca2+ indicator indo-1, but does not affect the ratio (R) of indo-1 fluorescence at 400 and 500 nm in the presence of caffeine concentrations up to 10 mM [1]. We have found that when caffeine is at concentrations of 20 mM or greater in vitro, or in saponinpermeabilized skeletal muscle fibers, a Ca(2+)-independent increase in R occurs, which leads to an overestimation of the free Ca2+ concentration. Depending on experimental conditions, two factors contribute to the alteration in R in vitro. First, when indo-1 fluorescence is low, fluorescence by caffeine, at 400 nm, can be significant. A second, and more dramatic effect, is that quenching of indo-1 fluorescence by 20-50 mM caffeine is dissimilar at 400 and 500 nm. Quenching at 500 nm is not linear, with respect to the concentration of caffeine, and causes a Ca(2+)-independent increase in R, that occurs even when the fluorescence of caffeine is a small portion of total fluorescence. However, unlike R, the Ca2+ calibration constant of indo-1, KD beta, is unchanged in 50 mM caffeine. Therefore, an accurate quantitation of Ca2+ in the presence of even high concentrations of caffeine can be made in vitro by determining the Ca2+ calibration factors of indo-1 (RMIN and RMAX) for each caffeine concentration. These effects of concentrations of caffeine greater than 20 mM are not observed in intact cells loaded with the cell permeant form of indo-1 when caffeine is applied extracellularly. This suggests either that the concentration of caffeine within the cell does not reach that necessary to produce the effect, or that the effects of caffeine on the dye are modified by the environment within the cell.