On March 2022, the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) released the updated Fukushima Report　2020/21. Through this critical review, I identified serious problems. (1) The Report introduced the novel terminology ‘discernible’ that is equivalent to statistically enough power to detect increases in cancer. The Report explains that ‘“no discernible increase” did not equate to an absence of risk (Para. 213)’. In fact, for ‘females of ages in utero to five years at initial exposure, about 16 to 50 cases of thyroid cancer attributable to radiation could be inferred (snip) A statistical power analysis showed that an excess of 50 cases or less would be undetectable (Para. 222)’. However, UNSCEAR’s news releases and briefings do not explain the meaning of ‘discernible’, leading reporters and the general public to misinterpret ‘no discernible’ is ‘no risk’. That is a fatal failure in risk communication. (2) For thyroid cancers identified in Fukushima Ultrasound Examination (TUE), the Report concludes ‘the excess does not appear to be associated with radiation exposure, but rather a result of the application of highly sensitive ultrasound screening procedures (Para. 246)’. Because compared to Chernobyl, where a significant increase was observed among children under five years after four years of the accident, in Fukushima, a large part of thyroid cancer was detected in adolescents within three years. The report missed the context of TUEs in Chornobyl and Fukushima: TUE in Chernobyl started in the 1990s or after 4–5 years of the accident; in Fukushima, TUE started after a half year. Moreover, the literature review in the Report does not understand the limitations of the studies that reported insignificant relationships between radiation and thyroid cancer, suffered a lack of statistical power, and/or improper analysis. (3) For solid cancer (excluding melanoma and thyroid cancer), UNSCEAR’s statistical power analysis in Attachment A-23 indicated a lifetime increase likely to be detectable (LFR is 1.2% for 10-year-old girls subpopulation with a statistical power of 0.80). On the contrary, the report describes, ‘the levels of exposure of members of the public have been too low for the Committee to expect distinctive increases in the incidence of breast cancer or other solid cancers (Para. 247)’. A Critical review revealed severe problems in the UNSCEAR2020/21 Fukushima Report; thus the Report must be corrected or updated accordingly.