Objectives: Social isolation is an increasing concern in older community-dwelling adults. There is growing need to determine effective interventions addressing social isolation. This study aimed to determine whether a relationship exists between physical activity (recreational and/or household-based) and social isolation. An examination was conducted for whether group- or home-based falls prevention exercise was associated with social isolation. Methods: Cross-sectional analysis of telephone survey data was used to investigate relationships between physical activity, health, age, gender, living arrangements, ethnicity and participation in group- or home-based falls prevention exercise on social isolation. Univariable and multivariable ordered logistic regression analyses were conducted. Results: Factors found to be significantly associated with reduced social isolation in multivariable analysis included living with a partner/spouse, reporting better general health, higher levels of household-based physical activity (OR = 1.03, CI = 1.01–1.05) and feeling less downhearted/depressed. Being more socially isolated was associated with symptoms of depression and a diagnosis of congestive heart failure (pseudo R2 = 0.104). Discussion: Findings suggest that household-based physical activity is related to social isolation in community-dwelling older adults. Further research is required to determine the nature of this relationship and to investigate the impact of group physical activity interventions on social isolation.