Long-term abnormal cervical curvature may lead to irreversible pathological changes in the cervical spine and cause serious paralysis. However, early abnormal cervical curvature can be corrected via cervical vertebra posture monitoring and therapeutic intervention. Therefore, for physicians, an effective monitoring system is necessary for detecting the patient's disease as early as possible. The main purpose of this study is to develop a wearable intelligent system to monitor cervical postures using a single tri-axis accelerometer. This system has a small size and lower power consumption can be easily worn for long-term. In order to evaluate the performance of the developed device, first, we used our prototype to classify seven different neck postures. The obtained results across five participants indicated that the average classification accuracy are 100% for each postures. Furthermore, our proposed device could distinguish four different cervical curvature levels according to pitch angles in the sagittal plane of body. Moreover, it could record the length of time during which neck posture is maintained at each cervical curvature levels, which is regarded as an important indicator for evaluating patients' neck status in clinical practice. Finally, we have developed a mobile application software and installed it in a smart phone to enable physicians monitor the status of patients' neck on-line with ease and achieve remote diagnosis.