In this paper the temperature control of a solar furnace is addressed. In particular, we propose the use of a feedback linearization generalized predictive control strategy where both the reference tracking task and the rejection of disturbances (represented by the variation of the input energy provided by the Sun, mainly because of the solar daily cycle and passing clouds) are considered. This allows the physical and security constraints to be explicitly taken into account in the design. Simulation and experimental results show the effectiveness of the methodology and that this kind of plants can be considered as a cheap or alternative option for the material treatment and testing in the industrial context. (C) 2013 Elsevier Ltd. All rights reserved.