The application relates to the technical field of through-wall radar, and discloses a method for estimating wall parameters in through-wall radar imaging. The method includes the following three stepsof training data collection, model establishment and model training, and specifically includes the following processes: firstly, obtaining a true value of wall parameters, and then carrying out sampling and constructing a data set, processing a signal obtained from a receiver, putting the processed signal data together with the true value of the wall parameters as the training data to form a training sample, establishing a training model based on polynomial regression, training the model by using the training data in the training sample to obtain an optimal model, and finally adopting the optimal model to solve the estimated value of the wall parameters. According to the scheme of the application, the characteristics of polynomial regression are used, the number of the highest terms is manually adjusted, the problem of over-fitting can be effectively avoided, and meanwhile, the model is small in calculation amount, and the requirements for timeliness can also be met.