Activation of neutrophils results in a one-electron reduction of oxygen to produce the superoxide anion and other oxygen-derived, microbicidal species. Evidence from many kinetic studies of oxygen-derived radicals generated by stimulated neutrophils in vitro shows that radical production is optimal at 37 degrees C but only lasts several minutes and then rapidly subsides. These findings support the widely held perception that the neutrophil's "oxidative burst" is a transitory event that peaks within minutes of stimulation and ends shortly thereafter. However, while some studies have shown that under controlled conditions stimulated neutrophils can generate superoxide continuously for several hours, others have observed that the superoxide formation by neutrophils stimulated in buffer at 37 degrees C does not persist. To reconcile the conflicting findings and to better understand neutrophil function, we have reinvestigated the effect of temperature on the kinetics of radical generation by PMA-stimulated cells. Electron paramagnetic resonance spectroscopy coupled with spin-trapping and SOD-inhibitable ferricytochrome c reduction were used to monitor superoxide production by neutrophils stimulated at either 25 degrees C or 37 degrees C in RPMI 1640 medium or in Hank's balanced salt solution. When oxygen was supplied continuously, neutrophils stimulated at 25 degrees C in buffer or in medium generated superoxide for several hours but at 37 degrees C, particularly in HBSS, O2- formation strikingly and rapidly decreased. This cessation of superoxide generation was reversible by lowering the temperature back to 25 degrees C. These data imply that in vivo neutrophils may be capable of generating oxy-radicals for prolonged periods. In part, our results may also explain the often observed termination of neutrophil-derived radical formation in vitro and help to dispel the perception that neutrophil-derived oxy-radical production is an ephemeral phenomenon.