Social media use has increased drastically over the past two decades among youth and adolescents. Accordingly, research on the potential impact of social media use on health outcomes for these age groups has become an important area of enquiry. Early evidence suggests that, while social media may provide pro-social benefits for youth, negative experiences associated with social media use (e.g. high levels of social comparison and cyberbullying) may have negative mental (i.e. depression, anxiety, suicide) and physical (i.e. sleep disturbances, poor diet, weight gain) health consequences. Despite the importance of these research questions, few studies have examined individual differences in the impact of social media on health, or leveraged conceptual models to conceptualise risk and resilience factors. Here, we review the existing literature on youth social media use and health outcomes, and argue that the allostatic load model addresses gaps in the knowledge base by characterising specific aspects of social media use relevant to health and by generating hypotheses concerning how social media use over time might influence health and well-being. We end by articulating future research directions, including examining causal mechanisms using longitudinal designs, studying these effects in minoritised groups, and working with key stakeholders (i.e. physicians, schoolteachers, counsellors, and nurses) to recognise common social-media-related health problems to identify points of prevention and intervention.