Inferences regarding the propagation mechanisms of dart leaders and return strokes are made from a comparison of the behavior of traveling waves on a lossy transmission line and the observed characteristics of these two lightning processes, in particular, their observed light profiles <Jordan and Uman, 1983; Jordan et al., 1997>. The bottom kilometer or so of the subsequent-stroke lightning channel is modeled as an R-L-C transmission line whose resistance per unit length R is first reduced by the downward propagating dart leader and then further reduced by the upward propagating return stroke, while variations in the inductance per unit length L and the capacitance per unit length C during the dart-leader and return-stroke processes are neglected. The transmission line is assumed to be linear (R=const) but different ahead of and behind either the dart-leader or the return-stroke front with any nonlinear effects occurring at the front. The resistance per unit length is estimated to be as follows: (1) ahead of the dart-leader front about 18 k&OHgr;/m or greater; (2) behind the dart-leader front (and ahead of the return-stroke front) about 3.5 &OHgr;/m; and (3) behind the return-stroke front about 0.035 &OHgr;/m. Comparison of the behavior of traveling waves on such a transmission line with the observed properties of luminosity waves associated with dart leaders and return strokes in conjunction with other observations suggests that there is a difference between these two lightning processes in terms of the dominant propagation mechanism. It appears that the return stroke is similar to a classical (linear) traveling wave that suffers appreciable attenuation and dispersion and whose advancement can be visualized as being due to the progressive discharging of elemental capacitors (previously charged by the leader process) of the equivalent R-L-C transmission line. Ionization does occur during the return-stroke process but has a relatively small effect on the wave propagation characteristics which are primarily determined by the transmission-line parameters ahead of the front, as opposed to being determined by the wave magnitude. On the other hand, the progression of the dart leader is apparently facilitated by sustained electrical breakdown at its front, so the downward propagation characteristics of the dart-leader wave are primarily determined by the wave magnitude, which largely determines the front electric field, as opposed to being determined by the transmission-line parameters of the channel ahead of the front. Thus the dart leader may be best described as a downward moving ionizing front which generates current waves that propagate upward along the dart-leader channel. Finally, it can be argued that the subsequent return stroke can be viewed as a ground reflection of the dart leader. ¿ 1998 American Geophysical Union