Monochlorobimane (MCB) has been used by several investigators as a fluorescent label for quantifying glutathione (GSH) levels in human peripheral blood mononuclear cells (PBMC). This paper describes a biochemical evaluation of this approach. PBMC were incubated with MCB (10-100 microM) and the fluorescence in extracellular medium and cell lysates was measured. Nonlinear curves were obtained in both cases and no "plateau" was reached. The majority of the fluorescence was in the medium. Gel permeation (Sephadex G-25) of the lysate indicated a linear increase in protein-bimane adduct formation, reaching about 50% of the intracellular fluorescence after 1 h. Fractionation of the deproteinized samples with Sephadex G-10 showed that only about one-third of the "low-molecular-weight" fluorescence could be ascribed to GSH-bimane, in either the lysate or the medium. Furthermore, about 40% of the free GSH in lysates appeared unbound even after 1 h of incubation. These data are in line with our observation of an extremely low activity in PBMCs of glutathione S-transferase under the conditions employed. Our findings indicate that many variables influence the cellular fluorescence, including the presence of alternative metabolic pathways for MCB and the rapid excretion of GSH-bimane out of the cell. This lack of specificity limits the value of MCB as a GSH probe for PBMC and confirms earlier suggestions that a careful biochemical evaluation is a prerequisite for its application to any particular cell type.