Objectives: This study identified the classes (i.e. patterns) of caregivers' activities, based on their engagements in caregiving activities, and explored the characteristics and the caregiver burden of these classes. Methods: This study was a secondary analysis of a cross-sectional survey on the profiles of family caregivers of older adults in Hong Kong. A latent class analysis approach was adopted to classify family caregivers (N = 932) according to their routine involvements in 17 daily caregiving activities: 6 activities of daily living (ADLs) and 8 instrumental activities of daily living activities (IADLs) in addition to emotional support, decision making, and financial support. Multinomial logistic regression and multiple linear regression illuminated the characteristics of the classes and compared their levels of caregiver burden. Results: The family caregivers fell into 5 classes: All-Round Care (High Demand, 19.5%), All-Round Care (Moderate Demand, 8.2%), Predominant IADLs Care (High Demand, 23.8%), Predominant IADLs Care (Moderate Demand, 32.5%), and Minimal ADLs and IADLs Care (Low Demand, 16.0%). These classes exhibited different characteristics in terms of care recipients' cognitive statuses and caregiver backgrounds. The levels of caregiver burden differed across classes; the All-Round Care (High Demand) class experienced the highest levels of caregiver burden. Discussion: This study contributes to existing scholarship by turning away from a predefined category of care tasks to explore the patterns of caregiving activities. By identifying caregiving activity patterns and understanding their associated characteristics and caregiver burden, prioritizing and targeting caregiver support interventions better is possible.