[en] In case of loss of coolant accident, the void reactivity of the CANDU fuel provides a positive reactivity and increases the reactor power abruptly. Therefore it is required to secure credibility of the void reactivity in the stage of nuclear design and analysis, which has motivated this study to assess the measurement data of CANDU fuel void reactivity. First of all, the measurement data of ZED-2 and DCA were reviewed and the recent results were used for the analysis of uncertainty and validation of reactor physics codes. In principle, both the ZED-2 and DCA experiments are performed to measure moderator critical height during the substitution process by experimental lattices. The overview of the experimental results indicated that the confidence level of the void reactivity is ± 10 %. The assessment of physics codes, POWDERPUFS-V and RFSP, has been performed using Phase-B measurement data of Wolsong nuclear power plant 2. However, because the void reactivity measurement data is not available from the Phase-B test, the measurement data of coolant/moderator temperature coefficient, which are believed to be close to the property of the void reactivity, were used to assess the performance of void reactivity calculation by the CANDU core physics codes. The comparison between the calculation and measurement has shown that the prediction errors of the coolant and moderator temperature coefficients are 11.9 % and 1.6 %, respectively, which are within the allowable design error of 25 %