Neighbourhood cohesion is increasingly recognised as a key determinant of health and health-related behaviours. Positive association between social support and physical activity have been demonstrated on an interpersonal level, there is less evidence at group-level. This study aimed to examine the association between neighbourhood cohesion and physical activity trajectories during the COVID-19 pandemic. Hypothesizing that higher neighbourhood cohesion was a protective factor against reduced physical activity during the pandemic. Data from Understand Society (wave 9, Jan 2017-Dec 2019), and the COVID-19 sub-study (waves 1, 5, 7) was used. Participants (N = 14,475) had baseline data and at least one Covid physical activity measure. We used linear mixed models with a random intercept and slope at the individual level and an unstructured correlation matrix to examine the association between neighbourhood cohesion and physical activity during the follow-up period. We found a significant reduction in physical activity (-441 MET-min/wk, (CI 374.51 - 507.65, p < 0.001) through the COVID-19 pandemic, and that higher neighbourhood cohesion was related to higher physical activity after control for covariates. There was a significant difference between neighbourhood cohesion categories and change seen in PA during the 39-month follow-up period (difference in change between lowest and highest neighbourhood cohesion categories = 373 MET-min/wk, p = 0.036), higher neighbourhood cohesion had a protective effect. Strong relationships between public health and urban planning sectors are needed to build communities with structures in place to support a sense of community, social interaction and attraction to the neighbourhood. This will help long-term neighbourhood cohesion and support increased physical activity.