In this paper, the synthesis of an algorithm for detecting a phase modulation disorder of received narrowband radio signals against a background of white Gaussian noise is performed. To analyze the effectiveness of the resulting algorithm, statistical modeling of the phase modulation disorder is carried out. The dependences of the probability of detection of a disorder on the signal-to-noise ratio and the length of the sequence of received signals at a fixed moment of disorder are obtained. It is shown that when a low-order phase modulation disorder is detected, the required values of the probability of correct detection are achieved at low signal-to-noise ratios, while with an increase in the modulation order and the use of its modifications, the efficiency indicators of the algorithm will deteriorate. The implementation of the algorithm in radio engineering applications will allow optimizing synchronization algorithms and reducing the amount of transmitted service information in telecommunication systems.