The goal of our study was to determine whether an apoptosis assay used after short-term drug exposure could predict selective toxicity to cancer cells. To this end we compared the effect of eight anticancer drugs and 10 toxic compounds without known antitumor activity in cultures of human breast cancer cells and normal diploid fibroblasts by Apoptosis ELISA and growth inhibition assays. There was an overlap in concentration values of drugs and toxins inhibiting proliferation in cancer cells. In contrast, Apoptosis ELISA clearly distinguished between the two groups of compounds. Anticancer drugs induced apoptosis in cancer cells at 0.0015-0.5 microM, while toxins were effective at much higher concentrations of 8.0-50.0 microM. Moreover, six out of the 10 toxins did not induce apoptosis in cancer cells. The normal:cancer cell (N:C) ratio for growth inhibiting concentrations was in a similar range for anticancer drugs and toxins. The N:C ratio for apoptosis inducing concentrations was 33-200 for anticancer drugs and 1.3-3.0 for toxins. Our data indicate that apoptosis assays could be used to detect selective toxicity of anticancer drugs by determining apoptosis induction in cancer cells or through a comparison of apoptosis-inducing concentrations in normal and cancer cells.