To limit the spread of COVID-19, most countries in the world have put in place measures which restrict mobility. The co-presence of several people in the same place of work, shopping, leisure or transport is considered a favourable vector for the transmission of the virus. However, this hypothesis remains to be verified in the light of the daily data available since the first wave of contamination. Does immobility reduce the spread of the COVID-19 pandemic? Does mobility contribute to the increase in the number of infections for all activities? This paper applies several pooled mean group–autoregressive distributed lag (PMG–ARDL) models to investigate the impact of immobility and daily mobility activities on the spread of the COVID-19 pandemic in European countries using daily data for the period from 12 March 2020 to 31 August 2021. The results of the PMG–ARDL models show that immobility and higher temperatures play a significant role in reducing the COVID-19 pandemic. The increase in mobility activities (grocery, retail, use of transit) is also positively associated with the number of new COVID-19 cases. The combined analysis with the Granger causality test shows that the relationship between mobility and COVID-19 goes in both directions, with the exception of grocery shopping, visits to parks and commuting mobility. The former favours the spread of COVID-19, while the next two have no causal relationship with COVID-19. The results confirm the role of immobility in mitigating the spread of the pandemic, but call into question the drastic policies of systematically closing all places of activity.