Abstract In their seminal work, Vanquickenborne and van Tiggelen proposed that the position of the lifted flame in a jet of fuel is that at which the turbulence characteristics give rise to a burning velocity on the stoichiometric contour equal to the incident mean velocity. Moreover, this burning velocity is just that of a premixed flame under the same turbulent conditions. Some objections to this approach are that it does not take account of the large eddy structures in the jet, nor of the intermittency of the turbulence in some regions where flame is stabilized. Recent imaging studies and direct numerical simulation (DNS) results indicate that in fact the flame propagates around the periphery of the large eddies. In the intermittent region, these are often too fuel-rich in their core to be flammable. The present study takes a large number of experimental results from the literature and uses empirical functions to describe the self-similarity laws for jets, including the intermittency, in order to estimate the turbulence and convection velocities associated with the large eddies. These parameters are computed on the time-averaged contour of maximum laminar burning velocity at the heights where the flames were observed to stabilize. Premixed turbulent burning velocities based on these conditions are then computed from recent correlations in the literature and compared with the convection velocities, which are the inferred turbulent burning velocities for this situation. It is found that, when the jet exit velocities are high enough for the flame to stabilize more than 20 jet diameters downstream, the ratio of the predicted to the experimentally inferred burning velocity is 1.7 ± 0.7. This is consistent with a premixed burning hypothesis for the large eddies if the random direction of propagation associated with the large eddies is recognized.