Antimicrobial efficacy, which is central to many aspects of medicine, is being rapidly eroded by bacterial resistance. Since new resistance can be induced by antimicrobial action, highly lethal agents that rapidly reduce bacterial burden during infection should help restrict the emergence of resistance. To improve lethal activity, recent work has focused on toxic reactive oxygen species (ROS) as part of the bactericidal activity of diverse antimicrobials. We report that when Escherichia coli was subjected to antimicrobial stress and the stressor was subsequently removed, both ROS accumulation and cell death continued to occur. Blocking ROS accumulation by exogenous mitigating agents slowed or inhibited poststressor death. Similar results were obtained with a temperature-sensitive mutational inhibition of DNA replication. Thus, bacteria exposed to lethal stressors may not die during treatment, as has long been thought; instead, death can occur after plating on drug-free agar due to poststress ROS-mediated toxicity. Examples are described in which (i) primary stress-mediated damage was insufficient to kill bacteria due to repair; (ii) ROS overcame repair (i.e., protection from anti-ROS agents was reduced by repair deficiencies); and (iii) killing was reduced by anti-oxidative stress genes acting before stress exposure. Enzymatic suppression of poststress ROS-mediated lethality by exogenous catalase supports a causal rather than a coincidental role for ROS in stress-mediated lethality, thereby countering challenges to ROS involvement in antimicrobial killing. We conclude that for a variety of stressors, lethal action derives, at least in part, from stimulation of a self-amplifying accumulation of ROS that overwhelms the repair of primary damage.