To the Editor We are writing to raise serious concerns about the Review article on radiofrequency radiation (RFR) and cancer by Dr Grimes published recently in JAMA Oncology. 1 Over the past decade, our research center has conducted many studies to investigate the potential adverse health effects of human exposure to radiofrequency electromagnetic fields, including the long-term effects, such as increased risk of cancer. Grimes claims that the current evidence from epidemiological data in humans and laboratory animals does not support a link between exposure to RFR and increased cancer risk. He also notes that methodological shortcomings have undermined much of the research conducted to date on this topic. The article by Grimes has several major shortcomings, including but not limited to the following.Grimes comments that epidemiological studies reject any link between RFR and cancer and states: "To date, no adverse health effects have been established as being caused by mobile phone use." 1 While animal studies have shown that RFR is able to increase levels of cellular reactive oxygen species (ROS), the concentration of ROS is not large enough to damage DNA at low exposure levels. 2 However, when the ROS level increases, for example, by increasing the exposure time, radiation power density or specific absorption rate, RFR can damage DNA. It should be noted that while some studies did not report a positive association between mobile phone use in the general population, their findings indicate an increased association between cancer risk and heavy mobile phone users (users with cumulative call time >400 hours). 3 Given this consideration, we have previously introduced a nonlinear J-shaped dose-response relationship for the potential carcinogenesis of RFR. This J-shaped dose-response relationship explains why it is unlikely to detect RFR-induced cancer at low exposure levels. This model also explains possible hormetic effects at low RFR exposure levels and irreversible damage and hence increased risk of cancer at high levels. 3 Moreover, we have addressed the methodological errors and shortcomings