The condition of the COVID-19 pandemic has resulted in the enactment of physical distancing and staying at home policy. With this isolation, face-to-face activities are replaced by online activities. Several studies have found that distancing has an impact on increasing anxiety and stress in college students. This is inseparable from their ability in social problem-solving. The aim of this study was to determine how loneliness, perceived stress, communication skills with peers influence social problem-solving. Participants were 702 freshmen from University X and domiciled in the Jakarta Greater Area and several other cities. The instruments were the Social Problem-Solving Inventory-Revised Short-Form, UCLA Loneliness Scale Short Version (ULS-8), Perceived Stress Scale, and the Interpersonal Communication Scale. The data was collected through an online questionnaire and were analyzed using hierarchical regression. The dynamics of research results show that loneliness has a significant effect on social problem-solving. When perceived stress is included in the model, the effect of loneliness on social problem-solving decreases and tends to be weak, although still significant, the effect of perceived stress is much stronger. When communication skills with peers are included in the model, the effect of loneliness is no longer significant and the effect of perceived stress decreases, but it becomes the strongest variable affecting social problem-solving. On the other side, the three independent variables have a great influence on solving social problems in freshmen.