Abstract: Benchmarking the experiments data of critical assemblies and formatting the benchmark data set in the requirements of International Criticality Safety Benchmark Evaluation will enrich the International Handbook of Evaluated Criticality Safety Benchmark Experiments, and can be used in the nuclear field for nuclear data verification, numerical software validation, reactor design and criticality analysis, et al. Firstly, the methodology with its foundation for the uncertainty analysis of experimental k_eff was described, and the requirements with its approaches for modeling critical assembly were introduced, then three configurations of critical assemblies (different diameter HEU metal annuli with internally reflected graphite cylinder) with their experimental k_eff results were introduced, lastly the uncertainties of experimental k_eff and the biases of modeling the experimental configurations were performed and the benchmarking results of experimental k_eff with their uncertainties were achieved. The results meet the requirements for the inclusion of the International Handbook of Evaluated Criticality Safety Benchmark Experiments.