Chapter 2 describes the multi-level cellular radiotherapy response models that are needed for the analysis and optimization of tumor response to radiotherapy, parameter estimation and development of new radiotherapy techniques. Special attention is paid to model validation by comparison with the results of clinical trials. The radiotherapy response models include a sum of several exponentials describing cell proliferation, cell killing, and cell clearance; therefore, the problem of data fitting is ill-posed and may produce large nonphysical fluctuations in reconstructed radiobiological parameters. We describe a regularization technique to overcome the problem of ill-posedness and discuss parameters estimation in radiotherapy for head-and-neck cancer.