Recent advancements in vehicular technology are expected to enhance traffic safety by either warning the drivers or by automating the tasks related to driving to reduce the human driver’s involvement. The driver warning systems (DWSs) are designed to warn drivers in unsafe situations such as forward collision, lane departure, or when changing lanes with vehicles in blind spot areas. These features only warn the driver but cannot perform the driving tasks. Advanced driver assistance systems (ADASs) can perform driving tasks such as accelerating, braking, and steering, thereby eliminating the role of the human driver in performing these tasks. However, ADASs currently require drivers to remain seated and regain control when the vehicle demands.A plethora of research is available on the operational and safety benefits of the DWSs and ADASs. Most of these studies focus on calibrating the driving behavior parameters to mimic vehicles with particular DWS or ADAS using microsimulation software or a driving simulator. Some researchers also performed field tests using vehicles equipped with DWSs or ADASs but in a controlled environment. The efficiency of DWSs or ADASs tested in laboratories or controlled environments may vary depending on driving conditions and the complexity of driving tasks, demanding research on the factors affecting crash occurrence when driving such vehicles in normal driving conditions with other vehicles. Existing literature documents numerous studies focused on identifying factors affecting fatal crashes. The studies on fatal crashes show that factors such as roadway geometry, traffic control devices, vehicular characteristics, and other on- and off-road characteristics affect fatal crash occurrence. The factors related to the driver, such as attentiveness, distraction, and fatigue, also affect the crash occurrence. Although the DWSs and ADASs are designed to enhance safety, recent crash data shows that vehicles equipped with these systems still get involved in crashes. The reason for the same has been identified as either disengagement of the features or the risk other drivers possess to the drivers of vehicles with DWSs or ADASs. In addition, change in drivers’ behavior due to ADASs is also one of the factors influencing crash occurrence. The existing literature shows a dearth of research conducted to identify the factors influencing fatal crashes and fatal crash occurrence, considering the real-world crash data of vehicles equipped with varying DWSs and ADASs. Therefore, a comprehensive analysis considering the reported fatal crash data is imperative as it will help identify how the factors affecting fatal crash occurrence vary depending on the number and type of DWSs or ADASs equipped in the vehicles. In addition, conducting a study using a particular DWS or ADAS and the corresponding crash type for which the particular feature is designed would provide insights into the overall effectiveness of these features in terms of traffic safety. The findings from the study will assist in improving safety and proactively planning for infrastructure at higher penetration of vehicles with DWSs or ADASs. The objectives of the research, therefore, are (1) to collect and comprehensively evaluate data pertaining to the vehicles equipped with individual DWS and ADAS, (2) to identify, model, and compare factors affecting fatal crashes involving vehicles with individual DWS, and ADAS, (3) to identify, model, and compare factors affecting fatal crashes involving vehicles with one or more DWSs and ADASs with vehicles without any warning or assistance systems, and (4) to examine the effect of traffic characteristics, geometric characteristics, on-network, and off-network characteristics, and vehicle characteristics on the safety of vehicles with varying numbers of DWSs and ADASs present in the vehicles involved in fatal crashes. The fatal crash data is used to accomplish these objectives. The fatal crash data contains Vehicle Identification Numbers (VINs) of all vehicles involved in crashes, with information about all DWSs and ADASs equipped in the vehicles. In addition, the fatal crash data is more detailed and accurate than other crash severity data. Therefore, using only fatal crashes would provide reliable model estimates. The fatal crash data from 2016 to 2020 were obtained from the Fatality Analysis and Reporting System (FARS) database and considered for analysis and modeling in this study. Using VINs of all vehicles involved in crashes, information about vehicular characteristics such as type and the number of DWSs or ADASs was retrieved from the National Highway Traffic Safety Administration (NHTSA) database. The vehicular information is combined with fatal crash data to classify vehicles based on various DWSs or ADASs equipped in vehicles. In this study, DWSs such as Forward Collision Warning System (FCWS), Blind Spot Monitoring (BSM), Lane Departure Warning (LDW), and ADASs such as Lane-Keeping Assist (LKA), Adaptive Cruise Control (ACC), and Pedestrian Automatic Emergency Braking (PAEB) system were considered for the analysis. The combined dataset was further divided into three separate datasets, (a) multivehicle crashes, (b) single-vehicle and lane departure-related crashes, and (c) pedestrian crashes, to facilitate the analysis for different DWSs and ADASs depending on the type of crashes for which they are designed to enhance safety. A descriptive analysis of divided datasets was conducted, which showed that the proportion of crashes involving vehicles with DWSs or ADASs was less than 3% of the entire dataset. The locations of crashes were mapped to identify the spatial variation of crashes involving vehicles with various DWSs and ADASs. The temporal trends in the number of crashes involving vehicles with DWSs and ADASs were also plotted. The data visualization results showed that crashes involving vehicles with a particular DWS or ADAS vary spatially and temporally. Therefore, a comprehensive methodological framework to incorporate unobserved heterogeneity due to varying spatial, temporal, and driving behavior characteristics is proposed. The aspect of heterogeneity was addressed in three parts. Nearest neighbor analysis was conducted for each year of crash data of a particular dataset to account for spatial heterogeneity and sample crashes involving vehicles without a DWS or ADAS. The three nearest neighbors were obtained as the most optimal. The data from nearest neighbors and corresponding crashes involving vehicles with DWSs or ADASs were considered for modeling. The dependent variable in this study is either ordinal (number of DWSs or ADASs) or binary (with or without DWS or ADSA), depending upon the type of DWS and ADAS and the corresponding crash type for which a particular feature is designed. Logistic regression is the most appropriate modeling approach for these types of problems and was therefore used in the study. A fixed parameter and correlated random parameters ordered logit models were developed to identify factors affecting fatal crashes involving vehicles equipped with one or more DWSs and ADASs. In the case of pedestrian, single-vehicle and roadside departure-related crashes, a fixed and correlated binary logistic regression modeling approach was employed for vehicles with and without PAEB and LDW systems. The sole reason for developing random parameters model was to incorporate unobserved heterogeneity in modeling. To account for temporal heterogeneity, a temporal variable in the form of linear effect of time elapsed was included while modeling. Driver-related parameters in the dataset were considered as random parameters in correlated random parameters models to incorporate heterogeneity due to varying driving behavior. The goodness of fit indices such as Log-likelihood statistics and McFadden pseudo-r-square were used to compare and identify the best-fitted model amongst fixed and correlated random parameters models. Further, partial effects were obtained to derive inferences from the models. The results of the analysis conducted to identify factors affecting fatal crashes involving vehicles with and without DWSs or ADASs indicated that correlated random parameters models (ordered logit and binary logit) better fit the crash data. The correlated random parameters ordered logit model was significantly better compared to the fixed parameters ordered logit model. However, the difference in the goodness of fit indices was not statistically significant when the correlated random parameters binary logit model and fixed parameters binary logit model were compared, indicating that the improvement in model fit because of variation in driving behavior is not significant. The partial effects of models showed that vehicles with one or more DWSs or ADASs are more likely to get involved in fatal crashes in urban areas and on interstates. The probability of fatal crash occurrence for vehicles with LDW or PAEB during adverse weather conditions, such as ice, snow, smoke, or fog, was lower than for vehicles without those features. In wet or snowy road conditions, vehicles with DWSs, such as FCWS or BSM, and ADASs, such as LKA and ACC, are safer than vehicles without those features. However, vehicles with LDW and PAEB are unsafe during wet road surface conditions. On the other hand, vehicles with BSM, FCWS, LKA, or ACC are less likely to get involved in fatal crashes in conditions when the vehicle is skidding laterally or longitudinally before the crash. Similarly, the probability of fatal crash occurrence for vehicles with LDW is less when a vehicle is skidding longitudinally before the crash. In contrast, vehicles with PAEB are safer when the vehicle is skidding laterally before a crash. During critical road conditions, such as in the presence of work zones, vehicles with an ADAS or LDW are safer compared to vehicles without those features. In addition, vehicles with DWSs and ADASs, except those with PAEB, are safer at intersections than normal vehicles. In crashes related to speeding or driving under the influence of alcohol, vehicles with DWSs or ADASs are less likely to get involved in fatal multivehicle crashes. However, drivers traveling at a higher speed than the speed limit are more likely to get involved in fatal crashes in single-vehicle or lane departure-related and pedestrian-related crashes. From the results of all models, females and elderly drivers are more likely to get involved in fatal crashes when driving vehicles with any DWS or ADAS. In addition, it is notable that the probability of crash occurrence for vehicles with any DWSs or ADASs has increased from 2016 to 2020, showing that it is necessary to take precautionary measures to ensure better safety at higher penetration of these vehicles in the future. The data processing framework, methodological findings, and study results help identify the factors affecting fatal crashes involving vehicles with one or more DWSs or ADASs. The results of this study also highlight critical factors affecting fatal crash occurrence for vehicles equipped with individual or multiple DWSs and ADASs. The results help identify the potential areas for improvement in vehicular technologies for the industry. It also provides insights about factors related to road geometry and on-road and off-road characteristics to the practitioners, assisting them in better preparing the infrastructure for fully automated vehicles in the future.