In many situations, health care professionals need to evaluate the respiration rate (RR) for home patients. Moreover, when cases are more than health care providers’ capacity, it is important to follow up cases at home. In this paper, we present a complete system that enables healthcare providers to follow up with patients with respiratory-related diseases at home. The aim is to evaluate the use of a mobile phone’s accelerometer to capture respiration waveform from different patients using mobile phones. Whereas measurements are performed by patients themselves from home, and not by professional health care personnel, the signals captured by mobile phones are subjected to many unknowns. Therefore, the validity of the signals has to be evaluated first and before any processing. Proper signal processing algorithms can be used to prepare the captured waveform for RR computations. A validity check is considered at different stages using statistical measures and pathophysiological limitations. In this paper, a mobile application is developed to capture the accelerometer signals and send the data to a server at the health care facility. The server has a database of each patient’s signals considering patient privacy and security of information. All the validations and signal processing are performed on the server side. The patient’s condition can be followed up over a few days and an alarm system may be implemented at the server-side in case of respiration deterioration or when there is a risk of a patient’s need for hospitalization. The risk is determined based on respiration signal features extracted from the received respiration signal including RR, and Autoregressive (AR) moving average (ARMA) model parameters of the signal. Results showed that the presented method can be used at a larger scale enabling health care providers to monitor a large number of patients.