Studies examining the effect of social isolation on cognitive function typically involve older adults, and/or specialist groups (e.g., astronauts, Antarctic explorers). We considered the effects of COVID-19-induced long-term social isolation on cognitive function within a representative sample of the general population. We additionally considered how participants ‘shielding’ due to underlying health complications and participants living alone performed. We predicted that performance would be poorest under strictest, most-isolating conditions. At five timepoints over 13 weeks, participants (N=342; aged 18-72 years) completed online tasks measuring attention, memory, decision-making, time-estimation and learning, and indicated their mood, as ‘lockdown’ conditions were revised. Analyses revealed that performance typically improved as opportunities for social interaction increased. Interactions between participant sub-groups and timepoint demonstrated that performance was not shaped by repeated testing, but by individuals’ social isolation levels. Strategies to alleviate cognitive decline during prolonged restrictive conditions are required to avoid long-lasting impairment in the general population.