To analyze and manage the risk of drought in a region, access to some certain information is required, the most important of which is the frequency of drought severity-duration. Realizing the high correlation of these factors, one must pick up a method that pinpoint the relation and effects of these factors on drought analysis. In this paper, numerous copula families were employed to model the correlation structure of variables. Drought data and a three month standardized precipitation index in six stations of Tehran were employed. It is illustrated how to fit a model and compare different copula families to find the most appropriate copula function for the data. Next, by using the best selected copula function, the probabilistic behavior of drought from the severity–duration aspect was modeled based on the observations. The results show the relative closeness of frequencies of drought duration-severity in these stations, so that they can be presented for the province. In addition, the data were also analyzed based on the fitted model and the drought thresholds. That would make the results be more applicable, especially in bringing about the possibility of defining drought scenarios.