Decay heat calculations of spent nuclear fuel (SNF) using Polaris and ORIGEN codes in the SCALE code sys-tem, and CASMO5 code, are validated using measurements from the Clab and GE-Morris facilities. Multiple hypothesis testing, relying on permutations and bootstrapping, is conducted to simultaneously analyze the significance of differences between calculations and measurements - namely the biases. The biases are small for the Clab data, and larger and have more variance for the GE-Morris data. Then, sta-tistical tests are applied separately based on the measurement laboratory and the SNF origin. The calcu-lations on the Clab PWRs show to be systematically different from the measurements. The remaining benchmarks show to be insignificantly different from the measurements. The benchmarks are useful to validate decay heat calculations of PWR and BWR fuel assemblies at various cooling times. Nevertheless, the choice of the data and the codes could affect the significance of the obtained biases. (c) 2021 Published by Elsevier Ltd.