Rear-end collisions at unsignalized intersections located along high-speed intercity highways are one of the primary causes of crash fatalities. In recent years, traffic conflict technique (TCT)-based proactive safety evaluation utilizing proximal safety indicators for different conflict types has gained wide popularity. However, in developing countries like India, using readily available threshold values for these proximal indicators to identify critical rear-end conflicts might lead to misleading safety conclusions due to heterogeneous and indiscipline traffic conditions. Consequently, this study used two proximal indicators: time to collision (TTC) and deceleration rate to avoid collision (DRAC) to identify rear-end conflicts along intersections’ approaches from two crash-prone unsignalized intersections located in the National Capital Region (NCR), India. Critical rear-end conflicts were further classified using thresholds for these proximal indicators determined by extreme value theory (EVT). Considering different vehicle types, two-wheelers (2Ws) and light motor vehicles (LMVs) were predominantly involved in critical rear-end conflicts. Study outcomes suggested that despite the lack of reliable crash data in developing nations like India, combining proximal safety indicators with EVT proactively identifies critical conflict scenarios that could lead to collisions, without relying solely on historical crash data.