Dispersion modeling is increasingly being used (as opposed to downwind sampling) as the primary method for determining potential downwind particulate concentration levels resulting from a given source’s release. Modeling is attractive as a regulatory tool because of its ease of use, low cost, and small time requirement. On-site sampling demands greater costs, more time and manpower, and more care, while yielding less concentration data. As a result, sampling is used only as a last resort, and even then, may be set aside in favor of modeling results. The majority of the dispersion models approved by the Environmental Protection Agency (EPA) are Gaussian (or normal distribution) based models. These models require source parameter inputs, such as location and dimensions of emitting points and on-site structures and types and emitting rate of pollutants, along with meteorological data (wind speed, wind direction, etc…). This paper deals with the meteorological data input, and its effect on the modeling results. Three variations of meteorological data input format are compared, sub-hourly (two-minute), hourly average, and “spot” recordings. In general, the sub-hourly data results in concentration predictions that are more representative of actual conditions and have a lesser degree of variation in the concentration predictions.