Improving road safety remains a critical challenge in developing countries, where rural highways often experience a disproportionately high rate of crashes. Traditional crash-based methods for identifying hazardous locations are limited due to the underreporting and unreliability of historical crash data. This study introduces a proactive approach to highway safety assessment through the use of surrogate safety measures, focusing on traffic conflicts as indicators of potential danger. A methodological framework was developed to detect and analyze conflicts using data gathered from selected intersections via floating vehicle surveys and field observations. The study employs a count data modeling approach specifically, Poisson regression with both fixed and random parameters to evaluate how various road, traffic, and environmental factors influence the occurrence of traffic conflicts. Post-Encroachment Time (PET) and Time-to-Collision (TTC) were used as key surrogate indicators to classify conflict severity. Conflict frequency was analyzed across road segments, and high-risk locations were prioritized based on their predicted safety levels. Results demonstrate that surrogate safety indicators can effectively identify conflict-prone areas, offering a cost-efficient and reliable alternative to crash data in resource-limited contexts. The model’s partial effect estimates provide valuable insights into the impact of traffic flow, geometric features, and pedestrian activity on conflict likelihood. This conflict-based method offers practical value for road agencies in developing countries, enabling them to prioritize safety interventions more efficiently and proactively, even in the absence of comprehensive crash records.