It has been observed that technological innovations like Geographic Information Systems (GIS), Machine Learning, Artificial Intelligence (AI), Internet of Things (IoT), Big Data, and Intelligent transportation systems can offer useful techniques for identifying and providing details on variables impacting road safety. This is because there is no chance of human error when using these technologies to gather and analyze data. The majority of research on road safety focuses on predicting and averting technological, organizational, and human errors that may cause serious issues or collisions. Traffic accidents are recognized as one of the major global causes of injury and death. Trucks, cars, buses, motorbikes, and pedestrians are involved in crashes that result in almost 3700 fatalities and over 1.3 million deaths. This article aims to examine how well Naive Bayes classifiers perform in the domain of traffic crash prediction for the deceased class. The main objective is to assess and contrast the effectiveness of the three Naive Bayes variations in correctly categorizing occurrences linked to fatalities in traffic accident incidents, which are Bernoulli Naive Bayes, Multinomial Naive Bayes, and Gaussian Naive Bayes. Our study’s findings show that the three Naive Bayes classifiers perform consistently better when classifying traffic accident outcomes.