Normal-incidence surface-induced dissociation of singly protonated des-Arg(1)-bradykinin (PPGFSPFR) and des-Arg(9)-bradykinin (RPPGFSPF) has been studied using a specially designed Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR MS). We found that, with a reaction time of 1 s, the collision-energy-resolved fragmentation efficiency curve (FEC) for des-Arg(9)-bradykinin is shifted to lower energies by about 4 eV relative to the FEC for des-Arg(1)-bradykinin. Because the Arrhenius activation energies found by Williams and co-workers in BIRD experiments are 0.82 and 1.2 eV for des-Arg(1)- and des-Arg(9)-bradykinin, respectively, we expected to find the reverse order in our long-reaction-time FTICR-SID experiments. We rationalize the difference between our data and the thermal kinetics using Tolman's theorem to calculate threshold energies from the Arrhenius activation parameters. We found that, when the preexponential factor for reaction is very high (>10(17) s(-1)) or very low (<10(10) s(-1)), Tolman's correction becomes fairly large. In this case, the Arrhenius parameters are strongly correlated, and the activation energy depends on the reaction dynamics and temperature. This can result in the reversal of the order of the Arrhenius activation energies for different systems relative to the order of the threshold energies, meaning that conclusions about relative stabilities of large molecules should be based not on the thermal activation energies but rather on actual threshold energies, which can be calculated using Tolman's correction. The threshold energies extracted from the Arrhenius activation parameters are 1.15 and 1.24 eV for des-Arg(1)- and des-Arg(9)-bradykinin, respectively, assuming a temperature of 450 K. However, des-Arg(1)-bradykinin dissociates via a very tight transition state. Consequently, microcanonical rate-energy dependencies for these two peptides cross at very low internal energies and des-Arg(9)-bradykinin has a larger rate constant at internal energies sampled experimentally.