In recent years, regional bus services have been facing serious challenges such as a decline in ridership and a shortage of drivers. To address these issues and develop efficient operation plans, the collection of origin–destination (OD) data is essential. However, conventional OD data collection methods suffer from several problems, including labor and time constraints, missing data for flat-rate and cash users when using IC card records, and privacy concerns associated with face recognition cameras. To overcome these limitations, this study focuses on the use of Bluetooth Low Energy (BLE) advertising packets emitted by passengers’ mobile devices. A major challenge, however, is that modern mobile devices frequently and randomly change their Media Access Control (MAC) addresses, making continuous device tracking throughout the entire ride difficult. To address this, we propose an OD estimation method that enables continuous device tracking by applying a MAC address carry-over technique, which leverages the asynchronous timing of MAC address changes across nearby devices and information contained in BLE packets. To evaluate the effectiveness of the proposed method, field experiments were conducted on two bus routes with different characteristics. The results show that the proposed method achieved an average precision of 0.458 and an average recall of 0.407. Compared with our previous work, the average recall increased from 0.347 to 0.407, whereas the average precision decreased from 0.702 to 0.458 and the F1-score decreased from 0.465 to 0.431. These results indicate that the proposed method can detect OD records that were missed by the prior method, but it also increases false positives, revealing a trade-off between recall and precision.