This repository contains the datasets, Python scripts, and supporting documentation used in the study: "Traffic Safety Forecasting, Policy Evaluation, and Reform-Based Projections in Jordan to 2030." The repository supports the complete analytical workflow used in the manuscript, including data preparation, descriptive analysis, seasonal naïve benchmarking, SARIMA modelling, Prophet forecasting, XGBoost forecasting, cross-scale validation, interrupted time series analysis (ITSA), sensitivity analysis, bootstrap confidence intervals, feature importance analysis, and policy-conditioned projections to 2030. Contents include: • Monthly traffic safety dataset for Jordan (1997–2024)• Annual traffic safety dataset and derived indicators (1985–2024)• Python scripts for all analyses• Reproducibility documentation and software requirements The analyses evaluate long-term traffic safety trends in Jordan, estimate the association between major legislative reforms and safety outcomes, and generate policy-conditioned future trajectories under alternative enforcement persistence scenarios. Keywords: traffic safety, road safety, Jordan, forecasting, SARIMA, Prophet, XGBoost, interrupted time series analysis, policy evaluation, transportation safety. Author: George SammourAffiliation: Princess Sumaya University for Technology, Amman, Jordan