PurposeLimited evidence exists regarding a group of nurses' physical activity patterns and association with resilience. Less is known about the physical activity health paradox in nurses (the positive health effects of leisure time physical activity vs the negative health effects of occupational physical activity). This study aimed to explore the profiles of intensive care nurses' physical activity behaviours and associations with resilience, following a developed study-specific job demands–recovery framework.Design/methodology/approachA cross-sectional study was conducted with intensive care unit (ICU) nurses to explore their physical activity profiles and associations with resilience. The Connor-Davidson Resilience Scale 25 (CD-RISC 25) was used to assess resilience, and accelerometry was utilised to record participants' four-day activity (two workdays, two non-workdays). Hierarchical cluster analysis was employed to define groups of nurses by activity behaviours.FindingsParticipants (N = 93) were classified as low actives (n = 19), standers (n = 36), sitters (n = 31) and movers (n = 7). During two 12-h shifts, movers had the highest mean level of dynamic standing and the lowest mean level of sitting. During two non-workdays, movers had the highest mean level of walking as well as the lowest mean level of sitting and sleep time.Originality/valueThe uniqueness of this study was that it analysed ICU nurses' physical activity profiles and associations with resilience using identified clusters. However, the small number of participants limited this study's ability to determine significant relationships between resilience and the grouped physical activity profiles.