This paper explores the effects of the lockdown measures on the population’s exercise patterns and aims to answer three questions: Were lockdown exercise patterns different from the ordinary patterns? What are the variables that explain the difference between ordinary habits and lockdown exercise frequency of participants? What are the correlates of the changes in post-lockdown exercise habits among people who exercised during lockdown? We estimated binary logits for participation and ordered logits to analyse frequency and the changes in sporting habits using the 2020 Spanish Survey of Sporting Habits. Our sample consisted of 5,154 observations, of which 3,060 people exercised at least once a month (59.4%) and 2,178 exercised during lockdown (42.3%). The results show that some determinants of the probability and frequency of exercising lost significance during lockdown. Moreover, women and higher-educated people exercised more frequently during lockdown –comparing to their ordinary habits– and women physically active during lockdown exercised less once mobility restrictions were lifted. According to our results, policy makers should focus their attention on men and lower-educated people in the event of future stay-at-home restrictions. Keywords: Exercise, participation, frequency, lockdown, COVID-19