Dahinden et al. reported that N-formylmethionyl-leucyl-phenylalanine (fMLP)-induced superoxide release from polymorphonuclear leukocytes (PMNs) lasted more than 60 min when the cells were allowed to attach to a petri dish before induction. In contrast, it lasted only for 2.5 min when cells were in suspension (J. Clin. Invest. 72: 113-121, 1983). In spite of this report, the effect of cell adhesion has been ignored in most spin trapping studies of superoxide release from PMNs. This study shows that most PMNs in a quartz flat electron paramagnetic resonance (EPR) cuvette which was placed horizontally adhered to the wall within 3 min. In contrast, if the cuvette was placed vertically, only 20-30% of the cells became adherent in 30 min. We performed spin trapping studies using 5,5-dimethylpyrroline-N-oxide (DMPO) as a spin trap, and monitored the effect of cell adhesion on superoxide generation. When spin trapping was conducted on PMNs in suspension, the EPR signal of superoxide adduct (DMPO-OOH) was undetectable after stimulation with fMLP. However, PMNs which were allowed to adhere to the cuvette after stimulation generated superoxide for hours. Moreover, when PMNs were allowed to adhere prior to the stimulation, the magnitude of superoxide release was augmented three-to fourfold. Unlike fMLP, phorbol myristate acetate (PMA), which has been most commonly used in spin trapping studies, induced superoxide release which was not influenced by cell adhesion. We emphasize the importance of specifying the cell-adhesion-state in spin trapping studies.