BackgroundConsiderable research has been devoted to examining the mental health conditions of patients with COVID-19 and medical staff attending to these patients during the COVID-19 pandemic. However, there are few insights concerning how the pandemic may take a toll on the mental health of the general population, and especially of nonpatients (ie, individuals who have not contracted COVID-19).ObjectiveThis study aimed to investigate the association between social media use and mental health conditions in the general population based on a national representative sample during the peak of the COVID-19 outbreak in China.MethodsWe formed a national representative sample (N=2185) comprising participants from 30 provinces across China, who were the first to experience the COVID-19 outbreak in the world. We administered a web-based survey to these participants to analyze social media use, health information support received via social media, and possible psychiatric disorders, including secondary traumatic stress (STS) and vicarious trauma (VT).ResultsSocial media use did not cause mental health issues, but it mediated the levels of traumatic emotions among nonpatients. Participants received health information support via social media, but excessive social media use led to elevated levels of stress (β=.175; P<.001), anxiety (β=.224; P<.001), depression (β=.201; P<.001), STS (β=.307; P<.001), and VT (β=.688; P<.001). Geographic location (or geolocation) and lockdown conditions also contributed to more instances of traumatic disorders. Participants living in big cities were more stressed than those living in rural areas (P=.02). Furthermore, participants from small cities or towns were more anxious (P=.01), stressed (P<.001), and depressed (P=.008) than those from rural areas. Obtaining more informational support (β=.165; P<.001) and emotional support (β=.144; P<.001) via social media increased their VT levels. Peer support received via social media increased both VT (β=.332; P<.001) and STS (β=.130; P<.001) levels. Moreover, geolocation moderated the relationships between emotional support on social media and VT (F2=3.549; P=.029) and the association between peer support and STS (F2=5.059; P=.006). Geolocation also interacted with health information support in predicting STS (F2=5.093; P=.006).ConclusionsCOVID-19 has taken a severe toll on the mental health of the general population, including individuals who have no history of psychiatric disorders or coronavirus infection. This study contributes to the literature by establishing the association between social media use and psychiatric disorders among the general public during the COVID-19 outbreak. The study findings suggest that the causes of such psychiatric disorders are complex and multifactorial, and social media use is a potential factor. The findings also highlight the experiences of people in China and can help global citizens and health policymakers to mitigate the effects of psychiatric disorders during this and other public health crises, which should be regarded as a key component of a global pandemic response.