Based on the perspective of terrorism, this study uses a bootstrap full-sample causality test and sub-sample rolling window estimations to explore the risk aversion capability and terrorist financing capability of gold's time-varying dimension. The results suggest that terrorist incidents (defined by the number of terrorist attacks) and terrorist brutality (measured by the number of fatalities produced by terrorist attacks) are important factors in explaining changes in gold prices, but their effects differ. With extensive media coverage, the cause of investors' panic has changed from terrorist incidents to terrorist brutality. The results from some sub-sample periods support the capital asset pricing model by showing that terrorist attacks push up gold prices through a “terror premium.” Considering reverse causality, gold price can be used as an early warning of terrorist attacks, but its terrorist financing function is not yet proven. In the context of increasingly serious terrorism, studying the interaction mechanism between terrorist attacks and gold prices can help investors reasonably plan investments and provide insights for the government on combatting terrorism and maintaining world peace.