Public mental health issues have gained growing attention from academics and policymakers due to their increasing prevalence and multiple adverse and severe consequences. Although some studies have supported the benefits of parks on mental health, the causal relationship between park accessibility and mental health remains unclear. By converting a large cross-sectional sample of 22,060 undergraduates nationwide in China into a quasi-panel dataset, this study untangled the causal impact of park accessibility on mental health benefits. We employed a quasi-experimental research design and used a difference-in-difference (DID) model to estimate the causal effects of park accessibility on depression symptoms within varying buffer sizes (i.e., 0.5 km, 1 km, 1.5 km, and 2 km). Furthermore, propensity score matching (PSM) and the Heckman selection model were employed to mitigate the selection bias caused by the prior differences of the treatment group and the control group. The results revealed that park accessibility had a positive effect on mental health and that its influence decreased with increased buffer sizes. Regarding the gender and living-cost differences, park accessibility within the 0.5 km and 1 km buffers had a greater mental health impact on females than on males, and it had a greater impact on high-living cost undergraduates than on low-living cost undergraduates. To increase the mental health benefits of undergraduate students, this study suggests that the provision of parks within a 1 km radius buffer surrounding the campus should be a priority to improve the mental health of undergraduates.