Traffic accidents remain a major public safety concern worldwide, yet the factors associated with accident severity are often examined in isolation or within limited temporal scopes. This study presents an exploratory statistical analysis of severe traffic accidents in the United States using the US Accidents (2016–2023) dataset, which contains millions of real-world traffic incident records collected from multiple heterogeneous sources. The analysis focuses on the relationship between accident severity and non-medical contextual factors, including temporal patterns (hour of day, day of week, seasonality, and long-term trends), environmental conditions (visibility and weather categories), infrastructural features (traffic signals and junctions), and spatial context (urban versus non-urban areas). Severe accidents are defined as events with a severity level of three or higher, reflecting substantial traffic impact. A combination of descriptive statistics, visualization techniques, and non-parametric statistical tests is employed, including chi-square tests for categorical associations and Mann–Whitney U tests for continuous variables. The results reveal pronounced temporal regularities, with higher probabilities of severe accidents during night-time hours and weekends, as well as clear seasonal and interannual variations. Reduced visibility and specific weather conditions are associated with significantly higher accident severity, while infrastructural and spatial factors further differentiate risk patterns. Additionally, accident duration is examined as a proxy measure of traffic disruption severity. All findings are interpreted as observational associations rather than causal relationships. The study emphasizes transparency regarding data limitations, reporting biases, and potential confounding factors. By providing a comprehensive, reproducible, and statistically grounded overview of severe traffic accident risk patterns, this work aims to contribute to exploratory research in traffic safety analysis and to support further hypothesis-driven investigations.