This study examined the correlation between depressive symptoms, interpersonal sensitivity, and social support before and during the COVID-19 pandemic and verified causal relationships among them. The study used Social Support Scale and Symptom Self-Rating Scale to investigate relevant variables. A total of 1,414 employees from company were recruited for this longitudinal study, which a follow up study was conducted on the same group of participants 1 year later. Paired sample t-test results showed that significant differences were only found in social support, not in depressive symptoms or interpersonal sensitivity. The results of correlation analysis showed that social support, depressive symptoms, and interpersonal sensitivity were significantly correlated between wave 1 and wave 2. The cross-lag autoregressive pathway showed that employees’ social support level, depressive symptoms, and interpersonal sensitivity all showed moderate stability. Crossing paths showed that wave 1 social support could significantly predict wave 2 depressive symptoms ( β = −0.21, p < 0.001) and wave 2 interpersonal sensitivity ( β = −0.21, p < 0.001). Wave 1 depressive symptoms ( β = −0.10, p < 0.01) could significantly predict wave 2 social support, while wave 1 interpersonal sensitivity ( β = 0.07, p = 0.10) could not predict wave 2 social support. Social support can be considered as a protective factor against mental health problems.