Purpose: This study focuses on homecoming traffic flow prediction for sustainable anticipation to overcome the surge in traffic flow on the Jakarta-Cikampek Toll Road. This study aims to identify traffic patterns based on historical data, develop a time-series prediction model (ARIMA), and evaluate congestion levels using the Volume Capacity Ratio (VCR). The main issue is the high traffic flow during homecoming, which requires predictive and proactive approaches. This study uses concepts and theories such as traffic management, traffic flow characteristics, and time series prediction models (ARIMA and decomposition). This quantitative study analyzed historical data from Jasa Marga (2019–2024). Research Methodology: This quantitative study analyzed historical data from Jasa Marga (2019–2024). Analytical techniques included in this study, such as stationarity tests, ARIMA parameter identification, and VCR calculations, were used to assess congestion. Results: The results indicate that peak homecoming traffic occurs from H-5 to H-1, whereas returning traffic peaks from H+2 to H+5. The SARIMA (1,1,2) (2,1,2)²² model was more accurate in capturing seasonal patterns than the decomposition model. The VCR indicator is more than 1.0 during peak days, which indicates a congested road. These findings support traffic management strategies, such as contraflow and one-way systems. In conclusion, historical data-based prediction models can be used to effectively anticipate future traffic congestion. Conclusions: Historical data models can effectively anticipate congestion and support contraflows in urban traffic. Limitations: Stakeholders must enhance their data, infrastructure, awareness, and sustainable transportation. Contributions: This study used the Jakarta–Cikampek Toll Road for the results.