The invention discloses a spiral inductor modeling method based on a random forest, and the method comprises the following steps: 1, obtaining structural parameters and corresponding performance parameters of a spiral inductor through ANSYS HFSS electromagnetic simulation software, and constructing an original data set of the spiral inductor; 2, performing normalization processing on the originaldata set of the spiral inductor; 3, randomly dividing the normalized original data set into an original training sample set and a test sample set; 4, training the training sample set by using a randomforest algorithm, and obtaining a trained spiral inductance prediction model by using structural parameters of the spiral inductance as input and performance parameters as output; and 5, inputting the test sample set into the trained spiral inductance prediction model to obtain corresponding spiral inductance performance parameters, and evaluating the prediction precision of the model. The spiralinductor modeling method based on the random forest is established, the development and optimization of spiral inductors by designers are simplified, the research efficiency is improved, the cost issaved, and the integrated circuit design is accelerated.