There is a worldwide interest in how lockdown affects physical activity (PA) during the COVID-19 pandemic. Although it has been shown that the mandated stay-at-home restrictions and self-isolation measures applied in different countries were accosiated with a reduction in physical exercise and activity, such results derive from studying only specific periods of lockdown. However, in order for this hypothesis to be tested, consecutive lockdown periods need to be examined separately. In this study we focus on PA change in Greek adults over time, during each of the last four weeks of lockdown in Greece. The web-based Active-Q questionnaire (see Supplementary file 1_Active-Q) was used to collect data prior to the COVID-19 crisis (PRE condition) and during lockdown measures (POST condition). The period of data collection (5 April to 3 May 2020) was divided into four phases (Ph-I, Ph-II, Ph-III, Ph-V), corresponding to the 3rd, 4th, 5th and 6th lockdown week respectively (out of a six-week total lockdown). There were four independent groups of respondents (G-I, G-II, G-III, G-V) who reported their age, weight, height and usual PA habits. Energy expenditure (EE) was calculated (MET-min/week; see Supplementary file 2_Data) in four main different domains (daily occupation activities, means of transportation to and from daily occupation, leisure time and regular sporting activities; see Supplementary file 3_Corresponding MET values). Each group's dataset corresponded to one of the aforementioned phases (G-I to Ph-I, G-II to Ph-II, and so on). Overall PA change (from PRE to POST condition) ranged from -21.50% in G-I (Ph-I) to -5.03 in G-V (Ph-V); PA change in male subgroups ranged from -26.10% in Ph-I to -13.64 in Ph-V; in female subgroups it ranged from -17.42% in Ph-I to -1.39 in Ph-V. Although the decline in overall PA is evident in all groups during each lockdown phase ( p <0.05), the combination of our data demonstrates that towards the end of lockdown this decline showed a gradual decreasing tendency.