The cytotoxic effects of short duration, high temperature, and long duration, low temperature hyperthermia were determined in human cells growing in culture. The human tumor cell lines A549 (lung carcinoma), WiDr (colon carcinoma), and U87MG (glioblastoma) were used. In addition, a normal human lung fibroblast cell type 18Lu was used. Sensitivity to direct cell killing was measured at 41, 43, and 45 degrees C. Heat induced perturbations of cell cycle and proliferation were also analyzed. The results obtained on sensitivity of the above human cell lines at 43 and 45 degrees C are similar to those of the previous work of others in that the human cell lines were observed to be relatively resistant to thermal killing at 43 or 45 degrees C, when compared to heat sensitive rodent cell lines. The comparison is important because most prior hyperthermic research has been performed with rodent cells and clinical protocols have been designed with the use of rodent data. In contrast to the 43 degrees C response, most of the human cells we tested were killed by 41 degrees C heating to an extent greater than that observed for rodent cells. The heat sensitivities of the four different human cell lines varied widely. This appeared to be due to differences in both intrinsic heat sensitivity and tolerance development. During 41 degrees C heating, human cells did not proliferate and cell cycle perturbations developed but did not correlate with sensitivity to killing. Our heat sensitivity measurements point out the shortcomings of using data derived from rodent systems to predict clinical outcome of hyperthermia therapy.