Anxiety and depression among university students were reported to be increased during the outbreak of CoVid-19. Various studies indicated that the phenomenon was triggered by the sudden enforcement of the studying-from-home (SFH) policy. Accordingly, we hypothesized that the students’ online learning efficacy (OLE) predicted the upsurge of the two mental health issues. However, SFH also elevated the students’ inclination towards social media (SM), and it brought further changes in some socio-psychological factors. We recruited 435 university students to respond to demographic items and the scales to measure factors such as perceived social support (PSS), societal mattering, and fear of missing out (FOMO) to test the hypothesis that OLE would no longer be a significant predictor of anxiety and depression when all the aforementioned variables are controlled for. The results of the hierarchical regression analyses supported our hypothesis on anxiety, while in predicting depression, OLE was still significant after controlling for the rest of the predictors. Societal mattering was the strongest protective factor against depression, whereas FOMO was the strongest risk factor of anxiety. Implications, limitations, and suggestions are discussed.