Spin trapping coupled with electron paramagnetic resonance (EPR) spectroscopy has surfaced as one of the most specific and reliable methods for identifying free radicals in biological systems. Despite extensive studies focused on the kinetics of radical trapping by cyclic nitrones, the mechanism has not been fully elucidated. Moreover, major controversies still persist even regarding the efficiency and the rate constants of the trapping reaction. The present research used pulse radiolysis for studying the reaction of 5,5-dimethyl-1-pyrroline N-oxide 1 (DMPO) and of the ester-containing derivative, 5-tert-butoxycarbonyl-5-methyl-1-pyrroline N-oxide 4, with (OH)-O-., O-2(.-), CO2.-, C-.(OH)(CH3)(2), (CH2OH)-C-., and (CH3)-C-.. The results reveal that radical trapping is far more complex then previously realized. Radiation chemical experiments combined with EPR demonstrate that about 30% of (OH)-O-. add to nitrones 1 and 4 at position 2, yielding the corresponding persistent aminoxyls. The remaining (OH)-O-. radicals form transient intermediates that rapidly decay bimolecularly. These transient intermediates react with oxygen with rate constants that are significantly lower than those generally reported for alkyl radicals, which suggests that they are not simple carbon-centered radicals generated as a result of H-abstraction from the methyl or methylene groups of the nitrones. It is also shown that the addition of O-2(.-) and various aliphatic radicals to the nitrones is an equilibrium process. The upper limit for the rate constant of the reaction of nitrone 4 with O-2(.-) was 3 M-1 s(-1). The rate constant for the reaction of nitrone 1 with O-2(.-) was determined to be 170 +/-40 M-1 s(-1). This value is significantly higher than those previously determined by following the formation of the corresponding aminoxyl by EPR, which indicates that the yield of the aminoxyl is only a small fraction of the reacting O-2(.-).