Objective Respiratory rate is an important physiological parameter to monitor in clinic practice. It is of great significance to monitor respiratory rate for the diagnosis and treatment of certain respiratory related diseases. Photoplethysmography contains much physiological information. This research aims to extract respiratory rate from photoplethysmography. Methods We extract and analyze envelop,interval,amplitude and area signals of photoplethysmography,and then we use cubic spline interpolation and frequency spectrum analysis to obtain respectively frequency. After that we adopt weighted-average method to get the final respiratory rate and compare the respiratory rates extracted by this method and the standard respiratory rates. Results There are good correlation in both time and frequency domain and there is a low error according to the statistical result of different data. Conclusions The method we presented can extract respiratory rate from photoplethysmography effectively and with a smaller amount of calculations,which may propose theory foundation for the realization of real-time,continuous and noninvasive measurement of respiratory rate.