Calcium-sensitive dual excitation dyes, such as fura-2, are now widely used to measure the free calcium concentration ([Ca2+]) in living cells. Preferentially, [Ca2+] is calculated in a ratiometric manner, but if calcium images need to be acquired at high temporal resolution, a potential drawback of ratiometry is that it requires equally fast switching of the excitation light between two wavelengths. To circumvent continuous excitation switching, some investigators have devised methods for calculating [Ca2+] from single-wavelength measurements combined with the acquisition of a single ratiometric pair of fluorescence images at the start of the recording. These methods, however, are based on the assumption that the concentration of the dye does not change during the experiment, a condition that is often not fulfilled. We describe here a method of single-wavelength calcium imaging, in which the dye concentration is estimated from ratiometric fluorescence image pairs acquired at regular intervals during the recording period, that furthermore includes a correction for the changing dye concentration in the calculation of [Ca2+].