We describe an on-the-fly kinetic study of gas-phase growth of multiwalled carbon nanotubes. The methodology employs electrical mobility classification of the CNT, which enables a direct measure of CNT length distribution in an aerosol reactor. The specific experiment employs two mobility classification steps. In the first step we mobility classify the catalyst particle, in this case Ni, created by pulsed laser ablation, to generate a stream of monodisperse particles. This then determined the diameter of the CNT, when a hydrocarbon/H2 mix is added in a heated aerosol reactor. A second electrical mobility classification step allows us to determine the length distribution of the CNTs. We found that CNT growth from ethylene required the addition of small amounts of water vapor, whereas growth from acetylene did not. We show that acetylene, which always has small amounts of acetone present when purchased, can provide the oxygen source to prevent catalyst coking. By varying the temperature of the growth, we were able to extract Arrhenius growth parameters. We found an activation energy for growth approximately 80 kJ mol(-1) from both acetylene and ethylene, which is considerably lower than previous works for substrate-grown CNTs (E(a) = 110-150 kJ mol(-1)). Furthermore, we observed that our aerosol CNT growth rates were about 2 orders of magnitude higher than those for substrate-grown CNTs. The dominant growth mechanism of CNT previously proposed is based upon bulk diffusion of carbon through nickel particles. However, on the basis of the lower activation energy found in this work, we proposed that the possible mechanism of gas-phase growth of CNT is correlated with both surface (E(a) = 29 kJ mol(-1)) and bulk diffusion (E(a) = 145 kJ mol(-1)) of carbon on nickel aerosol particles. Finally, the experimental approach described in this work should be amenable to other nanowire systems grown in the aerosol phase.