Fura-2 has become the most popular fluorescent probe with which to monitor dynamic changes in cytosolic free calcium in intact living cells. In this paper, we describe many of the currently recognized limitations to the use of Fura-2 in living cells and certain approaches which can circumvent some of these problems. Many of these problems are cell type specific, and include: (a) incomplete hydrolysis of Fura-2 acetoxymethyl ester bonds by cytosolic esterases, and the potential presence of either esterase resistant methyl ester complexes on the Fura-2/AM molecule or other as yet unidentified contaminants in commercial preparations of Fura-2/AM; (b) sequestration of Fura-2 in non-cytoplasmic compartments (i.e. cytoplasmic organelles); (c) dye loss (either active or passive) from labeled cells; (d) quenching of Fura-2 fluorescence by heavy metals; (e) photobleaching and photochemical formation of fluorescent non-Ca2+ sensitive Fura-2 species; (f) shifts in the absorption and emission spectra, as well as the Kd for Ca2+ of Fura-2 as a function of either polarity, viscosity, ionic strength or temperature of the probe environment; and (g) accurate calibration of the Fura-2 signal inside cells. Solutions to these problems include: (a) labeling of cells with Fura-2 pentapotassium salt (by scrape loading, microinjection or ATP permeabilization) to circumvent the problems of ester hydrolysis; (b) labeling of cells at low temperatures or after a 4 degrees C pre-chill to prevent intracellular organelle sequestration; (c) performance of experiments at lower than physiological temperatures (i.e. 15-33 degrees C) and use of ratio quantitation to remedy inaccuracies caused by dye leakage; (d) addition of N,N,N',N'-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) to chelate heavy metals; (e) use of low levels of excitation energy and high sensitivity detectors to minimize photobleaching or formation of fluorescent non-Ca2+ sensitive forms of Fura-2; and (f) the use of 340 nm and 365 nm (instead of 340 nm and 380 nm) for ratio imaging, which diminishes the potential contributions of artifacts of polarity, viscosity and ionic strength on calculated calcium concentrations, provides a measure of dye leakage from the cells, rate of Fura-2 photobleaching, and can be used to perform in situ calibration of Fura-2 fluorescence in intact cells; however, use of this wavelength pair diminishes the dynamic range of the ratio and thus makes it more sensitive to noise involved in photon detection. Failure to consider these potential problems may result in erroneous estimates of cytosolic free calcium.(ABSTRACT TRUNCATED AT 400 WORDS)