Summary We assess the feasibility of using the MTT assay as a measure of cell viability in chemosensitivity testing in ovarian malignancy. The assay utilises the conversion of the tetrazolium salt MTT to formazan by dehydrogenase enzymes in living cells. We show that the optical density of the formazan produced from MTT is directly proportional to the number of live cells tested. Optimum MTT (Hill, 1983;Carney & Winkler, 1985;Weisenthal & Lippman, 1985). Short term cultures lost favour with the introduction of clonogenic assays which measure the proliferative capacity of tumour stem cells. It was thought that the inhibition of colony formation by cytotoxic drugs was the best indicator of tumour sensitivity. However, interest in short term cultures has recently revived, as the value of assessing critical damage to essential cellular functions in both resting and proliferating cells is being recognised (Weisenthal & Lippman, 1985). Mosmann (1983) described a new approach to quantitate mitochondrial dehydrogenase activity first described more than 30 years ago (Black & Speer, 1954). The intervening development in semi-automated, microtitre techniques has made his method rapid and simple. The assay, which measures the conversion of a tetrazolium salt (MTT) to formazan, has already been successfully applied to drug screening in cell lines (Alley et al., 1988; Carmichael et al., 1987) and fresh leukaemic cells (Sargent & Taylor, 1989;Twentyman et al., 1989;Pieters et al., 1988 (1982). Once a cell suspension was obtained these samples were also separated by density gradient centrifugation and washed in HBSS before resuspending in culture medium as described above. Cell counts were performed using a haemocytometer and the concentration adjusted to 1 x 106 ml-'. The morphology of the cells was assessed prior to plating on a cytospin preparation. The number of malignant cells was also determined by immunocytochemistry using a standard APAAP technique with the monoclonal antibodies HMFG2 (Oxoid, Basingstoke) and Cam 5.2 (Becton-Dickinson, Cowley, Oxford). The viability of the cells was determined by trypan blue dye exclusion.The relationship of cell numbers to the absorbance of formazan produced was determined by incubation of a range of known cell concentrations with MTT for 4 h. The optimum concentration of MTT and the incubation period to allow adequate formazan production was also determined. Drug exposureEach sample obtained was incubated with up to six different cytotoxic drugs. Stock solutions were prepared in appropriate solvents to a concentration of 100 jig ml-I (cisplatin, doxorubicin, chlorambucil, mitoxantrone) and 1 mg mlh' (carboplatin, treosulfan) and stored at -20°C. The drugs were tested at four concentrations appropriate to the plasma levels achieved in vivo (Metcalfe, 1983). These were prepared in medium from stock solutions immediately before use. Cisplatin and chlorambucil were used in the range 0.625-5;Lgml-1; doxorubicin and mitoxantrone 0.1-1 fg ml-': treo-