As previous papers show, our group developed computer models simulating spatial (3D) tumor growth of an in vitro tumor spheroid. These models were extended by implementing irradiation models based on the linear-quadratic survival function in order to simulate and optimize radiation therapy schemes. The key idea in this study is to simulate different fractionation schemes (standard-, super-, hyperfractionation, irradiation with a weekly high dose) and to compare the model results with regard to their tumor effectiveness. After introducing simplified model assumptions the following treatment plans, as a result, represent an optimal scheduling: 1. the hyperfractionation (3 x 1...1.5 Gy per day) in the case of rapidly growing tumors; 2. the hyperfractionation (3 x 1...1.5 Gy per day) for moderately fast growing tumors; and 3. the treatment with a weekly high single dose (1 x 6 Gy per week) in the case of slowly growing tumors. The transfer of the results gained by simulating in vitro-experiments to clinical tumors are discussed. Single observations in clinical practice concerning the therapeutical benefit indicate a rather good agreement with the simulation results. Thereby, the possibility is given to comprehend in vitro tumor growth and clinical therapy schemes in a model and to successfully simulate optimal treatment schedules by computer experiments. This method enables a reduction of time-consuming studies prior to clinical therapy.