A stochastic model simulated the dispersal of the boll weevil,Anthonomus grandis grandisBoheman, from cotton fields. The model was developed to study various factors influencing the total distance traveled by individual weevils in a population. It describes several behavioral attributes of weevil flight, such as flight speed and duration, and selected atmospheric conditions such as wind speed and direction which affect flight. The dispersing weevil population is generated by a population dynamics model which is initialized with a population of weevils entering the field from overwintering sites. A rule-based expert system adjusts the number of overwintered weevils, larval mortality, insecticide spray threshold, and spray interval in the population model to determine the number of weevils available for flight each day. These values act as input to a flight subroutine, which produces a distance distribution for dispersing weevils from the field. Final displacement of each weevil is calculated by integrating information on random flight heading, speed, and duration given the direction and speed of the wind at different flight altitudes. Sensitivity analyses were performed on model parameters to determine their effect on dispersal distances.