Automobile manufacturers are trying to improve the interior noise environment in their cars.  A more thorough understanding of how people perceive the noise is an important step towards this goal.  The focus of the current research is on modeling the acceptability of time-varying wind noises containing gusts.  A listening test was designed containing sounds that were simulated on the computer, based on pre-defined airflow profiles.  The time-varying noises in the test follow one of two simple gusting scenarios.  The primary scenario contained two segments of steady wind flanking a series of consecutive equal-strength gusts.  The number, duration, and strength of the gusts were varied between sounds.  This was done to examine general trends of acceptability with modulation rate, modulation depth, and duration.  The second scenario contained two gusts of equal or unequal strength, occurring either without a break or separated by a time gap.  This was done to examine the relationship between people’s reactions to the individual gusts in a pattern and their reactions to the whole pattern.  A small number of steady-wind noises were included for reference.  Terms in an acceptability model containing a previously proposed gusting metric were estimated. Possible refinements to the metric and model are discussed.