Two new mathematical models (with and without thermal contact resistance) to describe the behavior of a particular thermal conductivity probe (TCP) have been developed and analyzed together with experimental data of agar and Ottawa sand (C109) obtained by Tamawski et al. (2009, 2011, 2013). One of the models has been studied in a previous companion paper, Part I (Dang and Leong, 2015) to theoretically investigate the accuracy of the model for predicting the TCP's temperature response and thermal conductivity of a sampling medium. It is found that, in order to reduce error in obtaining the thermal conductivity (km) of a sampling medium, the initially measured temperature response up to an appropriate time, to, must be disregarded in calculation; otherwise, error more than 10% may be resulted in determining km. The thermal contact resistance (TCR) at the interface between the TCP and the sampling medium is also investigated and realized to confirm with the literature. The thermal properties of the TCP's epoxy layer are required in the model for predicting the temperature response of the TCP; but it is found that their values play an insignificant effect in the obtained thermal conductivity of the sampling medium. In addition, if considered independently, the heating period of the TCP is found to produce more accurate km values than the cooling period of the TCP. (C) 2015 Elsevier Ltd. All rights reserved.