Counterfactual thinking refers to the mental representations of alternatives to past events. It plays an important role in changing behaviors and improving performance by converting information about past mistakes into plans for future action. As an important cognitive strategy, counterfactual thinking is used in behavioral intervention. The current study explores whether counterfactual thinking, which has a behavior-regulating function, improves the behavioral intention of pedestrians. Four frequent unsafe behaviors of pedestrians were chosen as the experimental materials. A one-factor between-subject study was designed. The participants were randomly divided into two groups, namely, experimental (counterfactual priming) and control (experience priming). To match the sample from the experimental and control groups, we excluded the effect of individual differences and measured the baseline of all participants based on factual thinking, that is, their intention to follow a traffic rule (rating on a 6-point Likert scale) and their reaction time in rating. Next, a modified sequential priming paradigm was applied to prime the counterfactual thinking (experimental condition) or past experience (control condition) of the participants. The participants were then asked about their intention to follow the traffic rule again. As dependent variables, both the behavioral intention and reaction time were recorded again. Results showed that the intention of all participants to follow traffic rules was positive. A paired-sample t-test was conducted to test the differences in intention and reaction times between the baseline and two treatments. The rating scores and reaction times in baseline and experimental conditions showed significant differences. Specifically, the behavioral intention of the experimental group was more positive and their reaction time in rating was faster than those in the baseline. However, the rating scores in the baseline and control conditions showed no significant differences, although the reaction time of the control group was faster than the baseline. Hence, the behavioral intention of the pedestrians generally relied on past experience, and counterfactual thinking had a positive effect on the behavioral intention of such pedestrians. To exclude the effects of individual differences, ANCOVA was used to explore the possible difference between the two groups, using the baseline rating score and reaction time as the covariates. After controlling for the rating score and reaction time of the baseline, post-test rating scores in these two groups were found to be insignificant. However, there existed obvious trend that the score of the counterfactual priming group was higher than that of the past experience group, and the reaction time of the former was significantly faster than that of the latter. Thus, counterfactual thinking enabled pedestrians to choose positive behavioral intention, and this facilitation effect was not only for the behavioral intention rating but also for the reaction time. In conclusion, the current study provides a better understanding of the improvement of behavioral intentions through counterfactual thinking. Furthermore, the behavioral intention of the pedestrians relies on past experience in daily life. However, it can be improved by priming through counterfactual thinking. In this way, counterfactual thinking can be an effective intervention strategy in traffic safety management.