The working cell temperature of high concentrator photovoltaic systems is a crucial parameter when analysing their performance and reliability. At the same time, due to the special features of this technology, the direct measurement of the cell temperature is very complex and is usually obtained by using different indirect methods. High concentrator photovoltaic modules in a system are operating at maximum power since they are connected to an inverter. So that, their cell temperature is lower than the cell temperature of a module at open-circuit voltage since an important part of the light power density is converted into electricity. In this paper, a procedure for indirectly estimating the cell temperature of a high concentrator photovoltaic system from atmospheric parameters is addressed. Therefore, this new procedure has the advantage that is valid for estimating the cell temperature of a system at any location of interest if the atmospheric parameters are available. To achieve this goal, two different methods are proposed: one based on simple mathematical relationships and another based on artificial intelligent techniques. Results show that both methods predicts the cell temperature of a module connected to an inverter with a low margin of error with a normalised root mean square error lower or equal than 3.3%, an absolute root mean square error lower or equal than 2 degrees C, a mean absolute error lower or equal then 1.5 degrees C, and a mean bias error and a mean relative error almost equal to 0%. (C) 2015 Elsevier Ltd. All rights reserved.