Background Social distancing measures (SDMs) are community-level interventions that aim to reduce person-toperson contacts in the community, thereby reducing the size and impact of an epidemic or pandemic to protect healthcare capacity. SDMs were a major part of the global response during the COVID-19 pandemic first to contain, then to mitigate, the spread of SARS-CoV-2 in the community. Common SDMs included limiting the size of gatherings, closing schools and/or workplaces, implementing workfrom-home or hybrid work arrangements, or more stringent restrictions on movement such as targeted or blanket lockdowns. In the United Kingdom, SDMs were first implemented in March 2020 with a stayat-home order issued by the end of that month. Subsequently, phased relaxing and tightening of SDMs were carried out as the pandemic evolved. This systematic review summarised the evidence for the effectiveness of nine SDMs around the world, including the United Kingdom. Challenges There were several challenges for this review. One was the large number of publications relating to COVID-19, with more than 50,000 titles screened in this study compared to about 5,000 for a moderatesized systematic review. Almost all of the studies included were observational in nature, which meant that there were intrinsic risks of bias that could have been avoided were conditions randomly assigned to study participants. There were no instances where only one form of SDM had been in place in a particular setting during the study period, making it challenging for studies to estimate the separate effect of each intervention. Key findings • The more stringent SDMs such as stay-at-home orders, restrictions on mass gatherings and closures were estimated to be most effective at reducing SARS-CoV-2 transmission. However, the attributable effect sizes varied across settings and time. • The review for stay-at-home orders included over half of the studies included in this review. 119 of 151 studies estimated that stay-home-orders were effective at reducing the impact of COVID19. • The evidence for school closures was mixed but 54 of 104 found a beneficial effect. Distancing measures such as spacing out student desks and limiting the capacity of classrooms were reported to be effective at reducing transmission in schools. • The effectiveness of restricting mass gatherings was mainly estimated using the Oxford COVID-19 Government Response Tracker data. Effect sizes generally increased as the stringency of restrictions increased, with some studies showing the largest effect when people were restricted to gathering in groups no larger than ten. • Most studies included in this review suggested that combinations of SDMs successfully slowed or even stopped SARS-CoV-2 transmission in the community. However, individual effects and optimal combinations of interventions, as well as the optimal timing to introduce and later relax particular measures, require further investigation. • This review did not include evidence for measures such as testing strategies or modes of travel and specifically focussed on SDMs, as these were examined elsewhere. Knowledge gaps Despite the large volume of literature dedicated to estimating the effectiveness of SDMs to reduce the transmission of SARS-CoV-2, the determination of intervention-specific effects and optimal combinations of interventions proved challenging with a rapidly evolving pandemic. Much of the current literature on the effects of SDMs identified in this review used readily available population-level data and employed statistical methods to disentangle the effects of interventions that were applied simultaneously. However, findings from the current evidence base and its known limitations can aid the development of a study protocol designed to disentangle these effects in anticipation of future largescale epidemics or pandemics, including influenza and potentially other coronavirus epidemics such as the Middle East Respiratory Syndrome. Meanwhile, further investigation into the optimal combinations and timing of SDMs to reduce SARS-CoV-2 transmission with minimal sociocultural impacts is a major knowledge gap that warrants further investigation. Further research could examine the drivers of policy impact such as the degree to which human behaviour changed because compliance with these measures, which can vary across time, populations, and settings, are key to the success of SDMs. The potential for interactions with individual protective measures is also worthy of further study. While lessons from SDMs implemented during the COVID-19 pandemic can be used to design future pandemic preparedness plans, it is likely that these plans should always be adapted to the transmission dynamics of new pandemic viruses to achieve the best outcomes for public health.