This study estimated possible impact of COVID-19 lockdown and relief measures, on severe food insecurity in Soshanguve. Binomial logit models, combine statistical and maximum likelihood features to estimate the odds of severe household food insecurity in the township. Predictably, the risk of severe food insecurity is significantly higher among low income households, with large families, the unemployed and those with lower education. As not all social grants received were spent on food consumption, social support weakly contributed to reduce household severe food insecurity impact of COVID-19. Hence, households receiving grants were not necessarily lifted out of food insecurity because not all of the grant was devoted to food consumption. This study contributes to the growing body of knowledge on how COVID-19 lockdown and relief measures impact on food security in poor neighbourhoods.