An increase in the intracellular generation of reactive oxygen species (ROS) such as superoxide (O2.-) and hydrogen peroxide (H2O2) or a defect in the anti-oxidant defense system renders the cells oxidatively stressed, a state harmful for cell survival. However, recent evidence has added a newer dimension to the effects of elevated intracellular ROS by demonstrating that a pro-oxidant state amplifies cell proliferation, either via direct stimulation of cell division and activation of transcription or indirectly by inhibiting the execution of the cell death signal. These findings seem to tie up well with the observations that certain cell types, in particular tumor cells, constitutively generate ROS that function as autocrine growth stimulation signals. We are working on the premise that sustained elevation of intracellular O2.- in tumor cells could contribute to drug resistance by inhibiting the execution of the death signal. Indeed, in our recent reports, we have shown that pharmacological or endogenous inhibition of intracellular O2.- production enhances tumor cell sensitivity to drug-induced apoptosis. On the contrary an increase in intracellular O2.- inhibits apoptosis via a direct or indirect effect on caspase activation pathways. We hypothesize that the divergent signaling by ROS is a function of their absolute intracellular concentration and the critical balance between O2.- and H2O2. It is when the cellular generation of O2.- and H2O2 is excessive, that cells succumb to this stress by either activating apoptosis, or undergo necrosis. The critical determinant between survival and apoptotic or necrotic cell death, we believe may be the cytosolic pH, downstream of ROS production. According to our model, survival is favored with a mild sustainable increase in intracellular O2.- that maintains cytosolic pH in the alkaline range. Apoptosis, on the other hand is a function of intracellular H2O2 production accompanied by reduction of the intracellular milieu, and more importantly a decrease in O2.- level and cytosolic acidification. Thus, in order to differentiate necrotic from apoptotic stress, we propose to refer to the mechanism of apoptosis induced by an increase in ROS production as "reductive stress" as opposed to the term "oxidative stress", which should appropriately be used for ROS–induced necrosis.