Abstract In the present work, the results of testing panel and box solar cookers are used to investigate the suitability of using the linear regression for estimation of the performance parameters of a solar cooker. The panel cooker and the box cooker were experimentally tested with glycerine and peanut oil, respectively. An exponential fitting to the plot of measured load temperature versus time is used to derive the linear regression between the instantaneous efficiency and the specific difference of temperature. The linear regression curve is compared with the corresponding experimental curve. Minor deviations are observed in the case of the panel cooker, but only in the middle part of the test. In the case of the box cooker, the deviations are very significant during the whole test. The present work presents the simplified formulation associated with the physical problem. It points out the importance of performing further research to develop a more accurate procedure. The determination of parameters based on the linear regression cannot be seen as a universal procedure applicable to all types of cookers. The opto-thermal ratio and the maximum achievable load temperature are overestimated.