In this work, we discuss three empirical models and introduce one more detailed model named YieldOpt. All models can be used to calculate the power output and energy yield of concentrating photovoltaic (CPV) modules under different ambient conditions. The YieldOpt model combines various modeling approaches: simple model of the atmospheric radiative transfer of sunshine for the spectral irradiance, a finite element method for thermal expansion, ray tracing for the optics, and a SPICE network model for the triple-junction solar cell. YieldOpt uses a number of constant and variable input parameters, for example, the external quantum efficiency of the cells, the temperature-dependent spectral optical efficiencies of the optics, the tracking accuracy, the direct normal irradiance, the aerosol optical depth, and the temperature of the lens and the solar cell. To verify the accuracy of the models, the I-V characteristics of five CPV modules have been measured in a 10-min interval over a period of 1year in Freiburg, Germany. Four modules equipped with industrial-standard lattice-matched triple-junction solar cells and one module equipped with metamorphic triple-junction solar cells are investigated. The higher accuracy of YieldOpt compared with the three empirical models in predicting the power output of all five CPV modules during this period is demonstrated. The energy yield over a period of 1year was predicted for all five CPV modules with a maximum deviation of 5% by the three empirical models and 3% by YieldOpt