Castor beans are one of the world's top ten oil crops, with approximately 50% castor oil and 50% castor meal being obtained after a two-stage pressing process (cold and hot pressing). Nevertheless, castor meal still contains a significant amount of residual oil, due to the mechanical limitations of current pressing technology, and this can be further extracted from by means of thermal pyrolysis. In the study, the Taguchi method is performed to optimize the thermal pyrolysis of castor meal in order to maximize the yield of pyrolytic oil. The effects of different parameters on castor meal pyrolysis are investigated, namely the pyrolytic temperature, residence time, heating rate and nitrogen flow rate. The pyrolysis control factors were selected and categorized into three levels. Based on the Taguchi design concept, an L-9 orthogonal array was chosen for the experiments. The results show that the maximum yield of 19.61% (g-pyrolytic oil/g-castor meal) pyrolytic oil is obtained when the castor meal is subjected to a pyrolytic temperature of 400 degrees C, residence time of 120 min, heating rate of 20 degrees C/min, and the nitrogen flow rate of 200 mL/min. The effective sequence of different parameters on castor meal pyrolysis is the nitrogen flow rate, pyrolytic temperature, residence time and heating rate, respectively. The results of the experiment confirm that there were some differences between the theoretical and experimental yields because of certain assumptions in the Taguchi method and the non-uniformity of castor meal. As to the external energy input for castor meal pyrolysis, a preliminary assessment of electric power consumption was 3.27 kW-hr for the treatment of 1 kg castor meal at 400 degrees C for a period of 60 min. (C) 2014 Elsevier Ltd. All rights reserved.