Pavement maintenance decision-making is receiving significant attention in recent research, since pavement infrastructure is aging and deteriorating. The decision-making process is mainly related to selecting the most appropriate maintenance intervention for pavement sections to ensure performance and enhance safety. Several preventive maintenance methods have been proposed in the previous studies, yet the potential of implementing Case-Based Reasoning (CBR) in pavement maintenance decision-making has been investigated rarely. The CBR is an artificial intelligence technique, it is knowledge-based on several known cases, which are used to adapt a solution for a new case through retrieving similar cases. This research introduces the CBR to the area of pavement management to select the most appropriate preventive maintenance strategy for flexible pavement sections. The needed database was extracted from maintenance cases at Long-Term Pavement Performance Program. The criteria used to characterize condition of each section were identified based on the common practices in pavement maintenance published in the literature and implemented in the field. To assign weights to the selected criteria, different machine learning techniques were tested, and subsequently, Random Forest (RF) algorithm was selected to be integrated with the proposed CBR method producing the CBR-RF framework. A case study was analyzed to validate the proposed framework and a sensitivity analysis was conducted to assess the influence of each criterion on case retrieval accuracy and overall framework performance. Results indicated that the CBR-RF approach could assist effectively in the preventive maintenance decision-making with regard to new cases by learning from the previous similar cases. Accordingly, several agencies can depend on the proposed framework, while facing similar decision-making problems. Future research can compare the CBR-RF framework with other machine learning algorithms using the same dataset included in this research.