A significant number of natural disasters have occurred in the State of Florida over the past years. At the disaster preparedness stage, the population is advised to relocate to some of the available emergency shelters. However, the evacuation process generally happens in an unorganized fashion, as evacuating individuals are not assigned to a specific emergency shelter. Furthermore, traveling along evacuation routes with a dense traffic flow is a quite challenging task especially for vulnerable population groups (including aging adults). This study proposed a set of statistical models for estimating various driving performance indicators (e.g., travel time, fatigue, lane deviation, break reaction time, crash frequency, minimum space headway) based on a large variety of relevant factors (e.g., traffic flow characteristics, roadway geometric characteristics, driver characteristics, weather conditions, temporary attributes) under emergency evacuation. The statistical models were developed based on the data, collected using a driving simulator and 115 participants with various socio-demographic characteristics (i.e., age, gender, occupation, income, marital status, health condition, driving experience under normal conditions and emergency evacuation, and others). The driving simulation scenarios were designed by varying the roadway geometric characteristics. The developed statistical models are expected to fill out the existing gaps in the state-of-the-art and the state-of-practice related to emergency evacuation of different population groups, especially vulnerable populations such as aging adults. Furthermore, findings from this research will facilitate the emergency evacuation process and ultimately will reduce the number of fatalities as a result of natural hazards.