Regional head elections (Pilkada) are a crucial instrument in Indonesia's democratic system. However, their implementation often faces transparency issues, such as allegations of fraud, money politics, and limited access to public information. To maintain public trust, election observers are needed whose performance can be objectively evaluated. This study developed a decision support system based on the Simple Multi-Attribute Rating Technique (SMART) method to evaluate the performance of election observers in North Sumatra Province. The SMART method was chosen because it can handle multi-criteria decision-making through a process of weighting, normalization, and ranking alternatives. Research data was obtained from documentation from the North Sumatra General Elections Commission (KPU) for the 2024 period, including criteria such as mobilization of state officials, intimidation, money politics, politicization of social assistance, violations of the quiet period, and findings at polling stations (TPS). The results indicate that the SMART-based evaluation system provides a more systematic, transparent, and accountable assessment mechanism. These findings are expected to provide policy recommendations for the KPU to improve the quality of monitoring and strengthen transparency in the implementation of regional elections.