The objective of this research is to model the crowd evacuation process providing dynamic spatial-temporal distribution information, which can minimize the human exposure risk (death or casualties) on the population implied by a specific evacuation policy when he or she was exposed to adverse effects under accidents. An open space evacuation model based on the stochastic Markov process was introduced here to estimate the spatial-temporal distribution of the evacuees during evacuation covering the estimation of affected areas, space discretization, nodes and links creation, etc. Then, according to the solution of the Markov process providing the expected distribution of the evacuees in the nodes of the area as a function of time and the dose-response relationship, the people's health effects (e.g., death, several kinds of injuries, etc.) suffered in the evacuation process can be calculated, so that the accident's health consequence can be determined. Finally, different emergency response policies can be evaluated with corresponding health consequence, so that the emergency policy can be optimized. (C) 2012 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.