Social Simulation Theory (SST) considers the function of dreaming to be the simulation of social events. The Sociality Bias and the Strengthening hypotheses of SST were tested. Social Content Scale (SCS) was developed to quantify social events. Additionally, we attempted to replicate a previous finding (McNamara et al., 2005, Psychological Science) of REM dreams as predisposed to aggressive , and NREM dreams to prosocial interactions. Further, we investigated the frequency and quality of interactions in late vs early REM and NREM dreams. Data consisted of wake, REM and NREM home dream reports (N = 232, 116, 116, respectively) from 15 students. Dreams overrepresented social events compared to wake reports, supporting the Sociality Bias hypothesis. However, the Strengthening Hypothesis was not supported. We weren't able to replicate the McNamara et al. finding, and no time of night effect was found. While SST gained partial support, further research on social contents in dreams is required .