Objectives: This study aimed at estimating the pre-pandemic and pandemic prevalence of loneliness and investigating the association of loneliness with social disconnectedness during social distancing strategies in the time of the COVID-19 pandemic period.Methods: We used data from the ELSI COVID-19 initiative with participants from the Brazilian Longitudinal Study of Aging (ELSI-Brazil), which comprised 4,431 participants aged 50 years and over. Loneliness (hardly ever/some of the time/often) was assessed by the question "In the past 30 days, how often did you feel alone/lonely?". Social disconnectedness included information on social contacts through virtual talking (i.e. telephone, Skype, WhatsApp, or social media) and outside-home meetings with people living in another household. Covariates included sociodemographic and health related characteristics. Multinomial logistic regression models were used to estimate odds ratios (OR) with their 95% confidence interval (CI).Results: The overall prevalence of loneliness during the pandemic was 23.9% (95% CI 20.7-27.5); lower than in the pre-pandemic period (32.8%; 95% CI 28.6-37.4). In the pandemic period, 20.1% (95% CI 16.9-23.6) reported some of the time feeling lonely and 3.9% (95% CI 3.1-4.8) reported often feeling lonely. In the fully adjusted model, virtual talking disconnectedness (OR=1.67; 95% CI 1.09-2.56) was positively associated with some of the time feeling lonely and outside-home disconnectedness (OR=0.33; 95% CI 0.18-0.60) was negatively associated with often feeling lonely.Conclusion: Individuals with virtual talking disconnectedness and without outside-home disconnectedness are at higher risk of loneliness during the time of COVID-19 pandemic. Stimulating virtual talking connectedness might have the potential to diminish loneliness despite steep outside-home disconnectedness.