Recent attempts to quantify the rate constant for radical capture in cytochrome P-450 hydroxylations employing substrates that are precursors to radicals that rearrange very rapidly have given widely differing apparent rate constants, suggesting that the consensus hydroxylation mechanism is incomplete or incorrect or that the probe substrates behave in an unexpected manner. We report cytochrome P-450 hydroxylations of a new, calibrated hypersensitive radical probe substrate, (trans,trans-2-tert-butoxy-3-phenylcyclopropyl)methane (1a), that permits discrimination between radical and cationic intermediates. Cytochrome P-450 oxidation of the methyl group in 1a gave unrearranged and rearranged hydroxylation products. Most of the rearranged products derived from a cationic intermediate apparently produced during the course of the hydroxylation reaction; this unanticipated process is the origin of the confusing results obtained with other probes. The radical species in the hydroxylation reaction has a lifetime of only 70 fs; it is not a true intermediate but part of a reacting ensemble. The small amount of radical rearrangement occurs because the insertion reaction is nonsynchronous with C-H bond cleavage leading C-O bond formation. The short radical lifetime also requires that the oxygen atom is within bonding distance of carbon at the instant of hydrogen abstraction; that is, a "side-on" approach of oxygen to the C-H bond is suggested as opposed to a linear C-H-O array of a conventional abstraction.