Fluorescein esters are employed in assays of cell viability, membrane permeability and esterase activity. The ester most widely used, fluorescein diacetate (FDA), has the disadvantage of rapid cellular efflux of its hydrolysis product fluorescein. This is particularly problematic for flow cytoenzymology (FCE), where fluorescence is measured in individual cells allowing identification of subpopulations differing in esterase activity and/or membrane characteristics. We present a comparison of FDA with two potentially improved substrate probes for FCE, carboxyfluorescein diacetate (CFDA) and bis(carboxyethyl)-carboxyfluorescein-tetra acetoxy methyl ester (BCECF-AM). Substrates were characterized in terms of reaction and product efflux kinetics in EMT6 mouse mammary tumour cells, together with inhibition kinetics for the carbamoylating agent BCNU. Intact viable cells were analysed by FCE and spectrofluorimetry, and the latter was also used for cell sonicates and purified esterase. CFDA and BCECF-AM enter cells and are hydrolysed more slowly than FDA. CFDA and FDA hydrolyses obey Michaelis-Menten kinetics with Km values of around 19 and 2 microM, respectively, whereas BCECF-AM hydrolysis deviates from this classical behaviour. BCNU (5 X 10(-4) M) inhibits FDA and BCECF-AM hydrolyses by approximately 50%, compared to 30% for CFDA. CFDA may be partly hydrolysed by membrane-bound esterases. Efflux half-lives were 16 min, 94 min and greater than 2 h for products of FDA, CFDA and BCECF-AM, respectively. We conclude that BCECF-AM is the optimal substrate probe for FCE. This study emphasizes the need to optimize various parameters when selecting a substrate for flow cytoenzymological assay or when loading other reporter fluorochromes into cells via lipophilic esters.