The radiation sensitivity of six human ovarian tumor cell lines was evaluated using sulforhodamine B (SRB), tetrazolium (MTT) and clonogenic assays. Radiobiological parameters calculated from a linear quadratic model (SF2, alpha, beta) as well as from a single-hit multitarget model (D(o), Dq, n) and from the area under the dose-response curve (mean inactivation dose; MID) were compared. If the values deduced from MTT experiments were statistically comparable to those obtained from clonogenic assays, significant differences were observed between SRB and the two other assays that concerned the results achieved with the highest radiation doses tested (6-8 Gy), yielding a surviving fraction of approximately 20%. In addition, the intra- and inter-experimental variation of SRB dramatically increased within this range of radiation doses. However, up to 6 Gy, the SRB assay proved to be statistically comparable to MTT and clonogenic assays, and allowed the calculation of SF2, alpha and MID radiobiological parameters.