In clinical practice, sensitivity of malignant cells to a given immunotoxin remains hypothetical, since standard test systems such as the protein synthesis inhibition assay or the cloning assay are not appropriate. This study evaluated the feasibility of a semi-routine procedure based on dye exclusion assay enumerating the percentage of living cells after fluorescein diacetate-propidium iodide staining. The validity of the method was evaluated using five different subclones derived from the CEM cell line, which expressed a wide range of sensitivity to T101 A-chain immunotoxin. The comparison between dye exclusion assay and standard test systems suggested that this method might allow an easy and reproducible semi-quantitative evaluation of the sensitivity of leukemia cells. In a series of 21 patients suffering from various blood diseases in which the malignant cells expressed the T65 antigen, dye exclusion assay could detect clear T101 immunotoxin cell sensitivity in about 50% of the cases. The mean density of T65 antigen on malignant cells was found to influence dramatically the sensitivity of target cells to T101 immunotoxin.