Objective After total knee arthroplasty (TKA), appropriate rehabilitation and physical activity at home is important for patients’ recovery. This study was designed to develop a system for remote monitoring the knee bending training and walking steps and distances for patients after TKA. Method The system consisted of data acquisition module with accelerometer, gyroscope, geomagnetic sensor, patient-side App that calculates knee bending angles and times, and walking steps and distances, cloud server, and doctor-side App that reads data from cloud server. The finite state machine method was used to calculate knee bending angle and times during sitting; adaptive zero crossing method was used to calculate walking steps and distances. Twelve healthy subjects (age 26.4±3.8 years、height 1.71m±0.05m、weight 63.6kg±8.5kg) and 12 TKA patients (age 62.1±6.6 years, height 1.65m±0.07m、weight 66.8kg±3.5kg) were recruited to verify the accuracy of the system measurement.Result The hardware and software were successfully developed for recording and transmitting knee bending and walking data to the cloud server through patient-side App, and doctors can read the data from the cloud server through the doctor-side App. The predictive accuracy of knee flexion times is 100%. The calculation errors of walking steps are 1.6±1.3% for healthy subjects and 3.0±2.2% for TKA patients, and the calculation error of the walking distance are 5.3±2.9% and 6.5±4.9% respectively.Conclusion The system developed in study can effectively record and transmit knee bending and walking data of patients, and provides an effective method for remote monitoring rehabilitation and physical activity at home for patients after TKA.