Conventional particle tracking techniques used to predict the dynamics and statistics of spray flows can be prohibitively expensive, requiring large computation times and significant data storage. Moreover, because of the discontinuous nature of the spray drops, data from a simulation of the flow do not produce smooth statistics unless the results from many simulations have been averaged. A new model has been presented previously that computes spray statistics directly, without simulating the flow, by closing a set of transport equations for the low-order moments of the droplet probability density function. The model is now extended to include nonlinear drag, heating, and vaporization effects. The Ranz-Marshall correlation and d(2) law are used for droplet heating and vaporization; however, any other correlation could be used without significant change to the overall spray model. Both nonvaporizing and vaporizing test cases with a quasi-one-dimensional flow geometry show very good agreement in comparison with a conventional Lagrangian simulation. With three notable exceptions, the vaporizing and nonvaporizing results are very similar.