How science popularization information affects public risk perception in controversial science and technology issues is an important concern shared by science popularization practice and risk communication research. Taking nuclear energy as an example, this study draws on the Extended Parallel Process Model (EPPM) and employs a 2×2×2 three-factor between-subjects survey experiment to examine the effects of uncertainty information, risk-benefit information, and risk-coping information on public risk perception. The results show that uncertainty information significantly increases public threat perception, whereas the independent effects of risk-benefit information and risk-coping information are not significant. Negative interactions are observed among different science popularization information strategies: both uncertainty information and risk-benefit information weaken the positive effect of coping information on risk coping perception. System trust positively moderates the effect of coping information. These findings suggest that science popularization on controversial science and technology issues should move beyond the linear assumption that "more information is better" and shift toward more targeted designs of information combinations, while treating the cultivation of system trust as a foundational condition for improving science popularization effectiveness. Theoretically, this study strengthens the applicability of the EPPM to controversial science popularization and further specifies its underlying mechanisms. Practically, it provides empirical evidence for science popularization on high-risk issues such as nuclear energy.