The estimation of weather parameters such as attenuation and rainfall rates from remotely sensed weather radar data has been based mainly on deterministic regression models. This letter introduces a new Gaussian mixture parameter estimator (GMPE)-based framework to incorporate prior knowledge into this process. The GMPE makes possible a versatile model for parameter estimation under all conditions without compromising accuracy. Observations from dual-polarized and dual-frequency radar sensors can be utilized in the GMPE in a very flexible manner. Simulation examples have demonstrated that the GMPE has better estimation error performance than traditional methods for parameter estimation applications, particularly for noisy observations. The impacts of mixture number and state vector selections in the GMPE are also discussed.