Although Enskog kinetic theory describes intermolecular collisional dynamics in terms of hard core molecular cross sections with inclusion of environmental crowding (excluded volume), more detailed analyses must incorporate the effects due to softness of the intermolecular potential and correlations of successive collisions. We present a simple model for the probability, p(r)(t), for a collision (defined as a velocity reversal) to occur at a time t after the last collision, that includes all the effects Listed. Although the model incorporates correlated collisions, it does so only in the sense that such correlations affect p(r)(t), an intrinsically Markovian property since it carries memory only back to the last collision. We also examine the sign velocity autocorrelation function in order to identify the strictly Markovian contributions to the collisional dynamics. (C) 1997 American Institute of Physics.