The invention discloses an electromagnetic environment identification method suitable for an external radiation source radar. The method comprises the steps of receiving radar echo data through an external radiation source radar to obtain amplitude data and power spectrum data; respectively carrying out preprocessing operation on the two types of data; performing statistical characteristic value extraction on the amplitude data, and performing length cutting on the power spectrum data; inputting the obtained amplitude characteristic value data into a trained and optimized amplitude model neural network classifier to predict and obtain an amplitude model category, and inputting the obtained power spectrum cutting data into a trained and optimized power spectrum model neural network classifier to predict and obtain a power spectrum model category; and returning electromagnetic environment category data in the electromagnetic environment joint decision matrix according to the predicted amplitude model category and the predicted power spectrum model category. By means of the mode, model identification can be effectively conducted on the electromagnetic environment where the external radiation source radar is located at present.