We have heated Chinese hamster ovary (CHO) cells in G1 phase, and followed them by microscopic observation for 8 days. Non-clonogenic cells were monitored with regard to when they died, and whether death was preceded by cell division. Such cells underwent either a division-independent or division-associated mode of death. At high survival levels most cells died a division-associated death, while at low survival levels division-independent death predominated. Mouse 10T1/2 cells died primarily according to the division-independent mode, regardless of survival level. These data are consistent with two distinct targets for the thermal killing of G1 cells, one of which involves the cell's ability to divide properly. The heat-induced division delay was similar for both clonogenic and non-clonogenic cells, suggesting that cell death was not due to lengthened cell cycle times. Hydrogen peroxide caused more division-independent death than heat, at isosurvival. The mode of death following cis-platinum treatment was largely division-associated, regardless of survival level. The combination of heat and cis-platinum caused synergistic killing that was due to a potentiation of cis-platinum killing by heat, since the pattern of death was clearly similar to that caused by cis-platinum alone.