The main objective of this study was to device-assess the levels of physical activity and sedentary behaviour patterns of older adults during the situation prior to the COVID-19 pandemic, home confinement, and phase-0 of the deescalation. We also aimed to analyse the effectiveness of an unsupervised home-based exercise routine to counteract the potential increase in sedentary behaviour during the periods within the pandemic. A total of 18 noninstitutionalized older adults( 78.4 ± 6.0 y.), members of the Spanish cohort of the EXERNET-Elder 3.0 project, participated in the study. They were recommended to perform an exercise prescription based on resistance, balance, and aerobic exercises during the pandemic. Wrist triaxial accelerometers (ActiGraph GT9X) were used to assess the percentage of sedentary time, physical activity, sedentary bouts and breaks of sedentary time. An ANOVA for repeated measures was performed to analyse the differences between the three different periods. During home quarantine, older adults spent more time in sedentary behaviours ( 71.6 ± 5.3 % ) in comparison with either the situation prior to the pandemic ( 65.5 ± 6.7 % ) or the ending of isolation ( 67.7 ± 7.1 % ) (all p < 0.05 ). Moreover, participants performed less bouts of physical activity and with a shorter duration during home quarantine (both p < 0.05 ). Additionally, no differences in the physical activity behaviours were found between the situation prior to the pandemic and the phase-0 of deescalation. According to our results, the home confinement could negatively affect health due to increased sedentary lifestyle and the reduction of physical activity. Therefore, our unsupervised exercise program does not seem to be a completely effective strategy at least in this period.