Dispersion and polydispersity of evaporating droplets in turbulent flows are investigated through a newly proposed stochastic model. The model is based on a first-order time series analysis, addresses the anisotropy of turbulence, and adequately takes into account the temporal correlations. It predicts temperature and vapor mass fraction fluctuations as well as velocity fluctuations. A set of averaged equations are derived for the carrier phase in homogeneous turbulence while the interaction between the carrier phase and the droplet phase are considered through source terms. The performance of the model is assessed by conducting simulations of droplet-laden homogeneous. shear flows. Good agreements are observed for various velocity, temperature and vapor mass fraction statistics with the DNS data. (C) 2004 Elsevier Ltd. All rights reserved.