The aim of this study was to explore associations between internet/email use in a large sample of older English adults with their social isolation and loneliness. Data from the English Longitudinal Study of Ageing Wave 8 were used, with complete data available for 4,492 men and women aged ⩾ 50 years (mean age = 64.3, standard deviation = 13.3; 51.7% males). Binomial logistic regression was used to analyse cross-sectional associations between internet/email use and social isolation and loneliness. The majority of older adults reported using the internet/email every day (69.3%), fewer participants reported once a week (8.5%), once a month (2.6%), once every three months (0.7%), less than every three months (1.5%) and never (17.4%). No significant associations were found between internet/email use and loneliness, however, non-linear associations were found for social isolation. Older adults using the internet/email either once a week (odds ratio (OR) = 0.60, 95% confidence interval (CI) = 0.49–0.72) or once a month (OR = 0.60, 95% CI = 0.45–0.80) were significantly less likely to be socially isolated than every day users; those using internet/email less than once every three months were significantly more likely to be socially isolated than every day users (OR = 2.87, 95% CI = 1.28–6.40). Once every three months and never users showed no difference in social isolation compared with every day users. Weak associations were found between different online activities and loneliness, and strong associations were found with social isolation. The study updated knowledge of older adults’ internet/email habits, devices used and activities engaged in online. Findings may be important for the design of digital behaviour change interventions in older adults, particularly in groups at risk of or interventions targeting loneliness and/or social isolation.